GB2628237A

GB2628237A - Care device for plant factory

Info

Publication number: GB2628237A
Application number: GB2405363.9A
Authority: GB
Inventors: Xu Yaliang; Yang Qichang; Li Qingming; Zheng Yinjian; Zheng Yi; Che Yuanpeng
Original assignee: Shenyang Siasun Robot and Automation Co Ltd; Institute of Urban Agriculture of Chinese Academy of Agricultural Sciences
Current assignee: Shenyang Siasun Robot and Automation Co Ltd; Institute of Urban Agriculture of Chinese Academy of Agricultural Sciences
Priority date: 2021-10-15
Filing date: 2022-10-14
Publication date: 2024-09-18
Anticipated expiration: 2042-10-14
Also published as: GB202405363D0; WO2023061483A1; GB2628237B

Abstract

A plant care device for a plant factory, at least comprising a shuttle trolley (1) capable of moving around a cultivation rack (8) in a planting space and a care and acquisition unit (2) disposed on the shuttle trolley (1) and capable of performing image acquisition of a plant planted on the cultivation rack (8), the care and acquisition unit (2) completing image acquisition of the plant placed on the cultivation rack (8) while following the movement of the shuttle trolley (1) around the cultivation rack (8) along a preset path; and a processing module (3) for marking the plants having abnormal growth states by comparing a plurality of plant images acquired by the care and acquisition unit (2) with each other, so that the shuttle trolley (1) can perform a secondary target point inspection operation according to the marking result, and acquire dual heterogeneous verification data for secondary verification of the marked plants by means of the care and acquisition unit (2) and a monitoring module (4) which are mounted on the shuttle trolley (1). By means of the configurations, the quality of the plants harvested in each batch can be counted and analyzed, traceability analysis is performed on quality detection information so as to find and summarize a more reasonable planting scheme of the plants, thereby improving the quality of the plant products produced by the plant factory.

Description

PLANT-CARING APPARATUS FOR PLANT FACTORY

BACKGROUND OF THE APPLICATION 5 1. Technical Field

[0001] The present invention generally relates to indoor plant cultivation, and more particularly to a plant-caring apparatus for a plant factory and a method thereof

2. Description of Related Art

[0002] In the modern agricultural production industry, factory fanning represents a kind of facility agriculture that integrates automation control and artificial intelligence and features high professionalization and high modernization. At present, factory farming mainly involves implementing technologies like soilless culture, nutrition control, CO2 fertilization, environment monitoring and control, and water-saving irrigation and using a smart control system so as to achieve efficient and modern plant production. With computer-based smart control and industrial fanning means, a plant factory can use preestablished nutrition control models and existing industrial cultivation technologies as well as experiences to realize plant cultivation and production. This is an industrialized method for effective and efficient factory production of plants.

[0003] Following greenhouse culture, plant factories have been developed and represent a kind of highly professionalized and modernized facility agriculture. Different from greenhouse production, a plant factory is totally freed from restrictions imposed by natural factors and climate to field production. Such a factory uses state-of-the-art devices to fully manually control environmental conditions, and thereby enables consistent supply of produce throughout a year. Currently, high-efficiency plant factories have grown fast in developed countries, and have preliminarily achieve factory production of vegetables, edible mushrooms, as well as precious flowers and trees. In the US, there are researchers using "plant factories" to grow wheat and paddy rice and run plant tissue culture, rapid propagation, and detoxification. Since the crop production environment in a plant factory is independent of external factors, such as climate, a high yield is ensured. For example, the first harvest of lettuce can be accomplished in two weeks after transplantation of the seedlings, and annually 20 or more harvests can be expected, which represents a yield equal to tens of times of that produced in an exposed field, and equal to at least 10 times as compared to a greenhouse culture. In addition, a plant factory supports soilless cultivation so as to produce pesticides-free, pollution-free vegetables. To date, there are only dozens of large-scale plant factories. This is because the high equipment investment and high power consumption (taking up more than a half of the total production costs) jointly lift the threshold that prevents extensive construction of plant factories around the world.

[0004] China Patent Publication No. CN102147127A discloses an air-conditioning system of a closed artificial light plant factory. The air-conditioning system comprises an artificial light source, a heat reclaiming device for the artificial light source, a timing controller and a heat reclamation air-conditioning unit. The timing controller controls on and off of the artificial light source. The artificial light source is turned off at daytime to form a dark period of plant growth, and the artificial light source is turned on at night to form a light period of plant growth. The heat reclaiming device for the artificial light source absorbs the heat of the artificial light source. Then the heat is absorbed by the heat reclamation air-conditioning unit. The heat is supplied to the plant factory through an air treatment unit. The heat reclamation air-conditioning unit and the air treatment unit are associatively controlled by the controller. The timing controller controls the artificial light source to be turned on and off in a way of day-night reversal, so that cold load at daytime is reduced, and cold load at nigh is increased. The utilization rate of the air-conditioning unit is improved, and the installed capacity is reduced, while the initial investment is reduced.

However, the prior-art system is merely a low-consumption circulation system for conditioning lighting and temperature in a plant factory and is unable to provide substantial care to plants or to automatically monitor plants for any abnormal growth condition.

[0005] China Patent Publication No. CN]00553443A discloses a closed, environmentally-controlled plant factory that solely utilizes artificial lighting. It comprises a building enclosure, an air circulation system, a temperature control system, a humidity control system, a CO2 supply system, and a light source system. It further comprises a control system designed with an embedded network and an automatic electricity-metering system. The factory is fully closed, heat insulated, and opaque, and depends merely on artificial lighting. The control system monitors and controls environmental factors in the factory, such as temperature, humidity, CO2 level, lighting, wind direction, and wind velocity, over a network structure, thereby combining air cleaning and environmental controls so as to provide a growth environment favorable to production of plants. The factory is claimed to have advantages about low costs and low energy consumption, and be directly applicable to mass rapid propagation and large-scale seedling production where plant genetic resources with no-pesticide farming, high quality and high adding values are desired, thereby realizing farming management and plan-based production with normalization and standardization. However, the prior-art technology is designed to monitor and control the general environment in the cultivation space as a whole, and is incapable of monitoring and controlling exact microenvironmental factors for any specific plant or autonomously performing patrol to detect whether there is any plant showing abnormal condition in its growth status.

[0006] Moreover, most existing plant factories lack the robust plant-caring equipment that can accurately monitor growth of all the plants in an extensive cultivation area without using considerable manpower, and this prevent them from early intervention for plants with abnormal growth conditions. Consequently, delayed discovery and intervention happening as late as in the late stage of cultivation may mean loss of the opportunity to timely rescue or improve plants from such abnormal development. Hence, there is a need for a plant-caring apparatus that regularly monitors plant growth so as to recognize plants with abnormal conditions and analyzes the abnormal conditions for proper treatment.

[0007] Since there may be discrepancy between the prior art comprehended by the applicant of this patent application and that known by the patent examiners and since there are many details and disclosures disclosed in literatures and patent documents that have been referred by the applicant during creation of the present invention not exhaustively recited here, it is to be noted that the present invention shall actually include technical features of all of these prior-art works, and the applicant reserves the right to supplement the application with the related art more existing technical features as support according to relevant regulations.

SUMMARY OF THE APPLICATION

[0008] In view of the shortcomings of the prior art, a technical scheme of the present invention provides a plant-caring apparatus for a plant factory. The apparatus comprises: a care-related collecting unit, being configured to monitor growth status of plants through collecting images of 50 the plants by means of infrared imaging and/or visual imaging; and a processing module, being configured to compare the images of the plants with each other so as to recognize and mark any said plant that exhibits an abnormal condition in the growth status, and control the care-related collecting unit and a monitoring module to collect double heterogeneous verification data for a follow-up confirmation operation performed on the marked plant, wherein the care-related collecting unit changes a working position thereof so as to shift itself among a plurality of observation angles around the marked plant while collecting the data for the follow-up confirmation operation, and the monitoring module changes a working position thereof in a way that it simulates working states of leaves of the marked plant so s to acquire data of microenvironment parameters that represent growth environments of the leaves.

[0009] A technical scheme of the present invention further provides a plant-caring apparatus for a plant factory. The apparatus at least comprises a shuttle cart that moves around a cultivation rack in a cultivation space and a care-related collecting unit that is mounted on the shuttle cart and captures images of plants planted on the cultivation rack. The care-related collecting unit, together with the shuttle cart, follows a predetermined route to move around the cultivation rack while capturing the images of the plants planted on the cultivation rack. A processing module compares the images of the plants captured by the care-related collecting unit with each other and marks the plants that exhibit abnormal conditions in the growth statuses, so that the shuttle cart can perform a secondary target-specific patrol and inspection according to the marking results, and use the care-related collecting unit and the monitoring module installed on the shuttle cart to collect double heterogeneous verification data for a follow-up confirmation operation performed on the marked plant. The present invention is advantageous for the following reasons. When a plant is suspected of growth abnormality, the present invention introduces a follow-up data collecting operation for further confirmation, wherein the caring apparatus automatically and accurately completes status determination of the suspect plant and preliminary analysis of the responsible disease. Additionally, the follow-up confirmation operation involves capturing clear images of the suspect plant at different angles for facilitating later diagnosis. Moreover, the monitoring module used herein is able to measure microenvironment parameters related to leaves of the monitored plants, thereby enabling analysis of data about the microenvironment around leaves of suspect plants and preliminary determination of whether the growth abnormality of the plant is caused by defective photosynthesis of the leaves due to environmental adverseness as a result of an unfavorable location of the plant on the cultivation rack. If it is the case, the adverse influences on the plant caused by nutrient solution supply in the cultivation rack and the current altitudinal location of the floating shelf carrying the plant with respect to the cultivation rack can be ascertained and corrected. Also, it is possible to determine whether there is any lighting module or another component on the cultivation rack broken according to the microenvironment parameters.

[0010] According to a preferred mode of the present invention, the care-related collecting unit collects the data for the follow-up confirmation operation in a way that it is driven by a displacement component to change working positions thereof so as to shift itself among a plurality of observation angles around the marked plant while collecting the images of the marked plant. The monitoring module changes a working position thereof in a way that it simulates working states of leaves of the marked plant, so as to acquire data of microenvironment parameters that represent growth environments of the leaves. The present invention is advantageous for the following reasons. By capturing clear images at different angles around the plant, the leaves or/and stems affected can be better identified, and this facilitates correct determination of the responsible disease later according to data in the sample library. Furthermore, the monitoring module that morphologically simulates leaves can collect data of microenvironment parameters right at the place the leaves are present, thereby enabling effective determination of whether the abnormal conditions of the plant are caused by unfavorable environmental factors. With this information, modules in the plant factory corresponding to the unfavorable environmental factors can be identified according to their abnormal data volume, so that targeted inspection can be performed for working states of these modules.

[0011] According to a preferred mode of the present invention, data of the images of the marked plant and the microenvironment parameters related to the leaves collected by the care-related collecting unit and the monitoring module, respectively, are simultaneously transmitted to a data-analyzing unit, which compares the data of the images of the marked plant against prestored parameter values from a sample database that represent average growth status of the plants in a growth cycle and determines whether the marked plant is wilting, lodging, chlorotic, or yellowing through analysis. The present invention is advantageous for the following reasons. The preload sample data can be directly used for comparison, and this reduces computing loads born by the data-analyzing unit. As the requirements for its operation is lowered, the data-analyzing unit can be economically realized using a low-cost processor.

[0012] According to a preferred mode of the present invention, the data-analyzing unit is further loaded with a sample library that contains reference values of the microenvironment parameters that enable the leaves to perform sufficient photosynthesis corresponding to various growth phases in the growth cycle of the plants, so that the data-analyzing unit compares the actual data of the microenvironment parameters related to the leaves of the marked plant collected by the monitoring module against the reference values related to the relevant growth phase and outputs analysis results.

[0013] According to a preferred mode of the present invention, upon marking of the suspicious plant, the processing module matches the images of the plants with coordinate points along a travel route of the shuttle cart, so that the images of the plants collected in a same time period can be used to determine respective locations of the plants on the cultivation rack through a reverse calculation and a coordinate location of the marked plant on the cultivation rack can be determined.

[0014] According to a preferred mode of the present invention, where analysis results output by the data-analyzing unit according to the data for the follow-up confirmation operation indicate existence of any condition seen in the marked plant that cannot be determined in situ whether it's a disease, the data-analyzing unit transmits data of the images of the marked plant to a control center of the plant factory for further disease interpretation and sample data storage.

[0015] The present invention further provides a plant-caring apparatus for a plant factory. The apparatus comprises a growth-facilitating device. The growth-facilitating device at least comprises a nutriment loop installed on a cultivation rack to supply nutrients to plants planted on all floating shelves of the cultivation rack through hydroponics or aeroponics. The plant-caring apparatus is characterized in comprising a data-analyzing unit that modulates the growth-facilitating device according to current physical states or characteristics of the plants as collected by a care-related collecting unit, so that the growth-facilitating device is enabled to variably supply lighting and nutrient according to growth status of the plants.

[0016] According to a preferred mode of the present invention, a shuttle cart is formed as a gantry-type structure to be able to span the nutriment loop and has two sides thereof each provided with a lifting unit to uphold a seedling tray placed on one said floating shelf [0017] According to a preferred mode of the present invention, the shuttle cart is provided with the care-related collecting unit and a robotic gripper for handling the plants, wherein the care-related collecting unit monitors the growth status of the plants, and the robotic gripper performs a planting, thinning or harvesting operation on the plants according to information related to the growth statuses.

[0018] According to a preferred mode of the present invention, after the shuttle cart enters a predetermined range for sensing a said plant, the shuttle cart collects current physical states or characteristics of the plant using the care-related collecting unit through infrared imaging and/or 10 visual imaging.

[0019] According to a preferred mode of the present invention, the nutriment loop comprises photosynthesis nutriment nozzles arranged at a top of each said floating shelf and rhizosphere nutriment nozzles arranged at a bottom of each said floating shelf wherein the photosynthesis nutriment nozzles release nutriments to stem and leaves of the plants in an aerosol-like form, and the rhizosphere nutriment nozzles release nutriments to roots of the plants in an aerosol-like form.

[0020] According to a preferred mode of the present invention, the shuttle cart is formed as a gantry-type structure to be able to span the nutriment loop and has two sides thereof each 20 provided with a lifting unit to uphold a seedling tray placed on one said floating shelf [0021] According to a preferred mode of the present invention, the shuttle cart is further configured to work with a loading/lifting machine to transfer cultivation plates on which the plants are planted.

[0022] According to a preferred mode of the present invention, the loading/lifting machine transfers the cultivation plates to a process unit, so that the process unit performs a processing operation on the plants planted on the cultivation plates

BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIG. 1 is a flowchart of operations performed by a plant-caring apparatus for a plant factory according to a preferred embodiment of the present invention; [0024] FIG. 2 is a flowchart of operations performed by a plant-caring apparatus for a plant factory as provided in Embodiment 4; and [0025] FIG. 3 is a flowchart of operations performed by a plant-caring apparatus for a plant factory as provided in Embodiment 5.

List of Reference Signs I shuttle cart 9: growth-facilitating device 2: care-related collecting unit 10: process unit 3: processing module 11: light supply unit 4: monitoring module 71: first robotic arm 5: data-analyzing unit 72: second robotic arm 6: control unit 300: plant growth monitoring device 7: displacement component 310: patrol cart 8: cultivation rack 320 data processing center

DETAILED DESCRIPTION OF THE APPLICATION

[0026] The present invention will be described in detail with reference to the accompanying drawings.

[0027] In a practical-use environment, data of physiological information of plants are collected using relevant devices, and nutrition supply is made to the plants according to setting of the nutrition control model and actual needs of the plants, thereby facilitating optimal growth of the plants. In order to achieve quality cultivation of plants in factory-based farming scenarios, plant factories have to monitor and modulate various environmental factors in a cultivation space that affect plants, such as the temperature, humidity, lighting, and the CO2 level, thereby creating and maintaining growth-facilitating conditions for plants. To this regard, data collection forms the basis of the monitoring and control process. For monitoring and modulating environmental factors, the prerequisite is to possess complete information of various environmental factors. A data-collecting system designed to accomplish such a data-collecting task comprises sensors as its essential parts. It is known that environmental factors are different from each other in nature and property, the data-collecting system has to employ at least temperature sensors, humidity sensors, lighting sensors, biological sensors, etc. The data collected by the data-collecting system are processed by a computer through statistical analysis and intelligent processing before displayed. A computer intelligence system, according to the displayed data and growth requirements of plants, controls the relevant systems and devices to operate and optimize various environmental factors, thereby performing plant production scientifically, orderly and manageably.

[0028] Currently, environmental control performed for factory farming of plants is mostly based on the assumption that atmospheric factors in a facility environment, such as temperature and humidity, are consistent throughout the whole space. However, the assumed consistency is not true and these factors in fact have spatially varying distributions. Hence, to precisely create optimal growth conditions for plants, it is necessary to control the spatially varying distributions of these influencing factors throughout the environment. Empirically, in a factory-type planting environment, the air humidity distribution correlates with temperature, velocity and direction of air flows in the indoor environment as well as the locations of plants with respect to air inlets. In a cultivation space of a real-world plant factory, distributions of air temperature and humidity around plants are mainly determined by temperature, velocity and direction of air flows in the cultivation space. Preferably, a neural network model is used to simulate and verify the relation between temperature/humidity distribution and the air flow characteristics in the given area, thereby ascertaining correlation between adjustable temperature/humidity distribution and the air flow characteristics in a cultivation space. With the knowledge, temperature distribution in a space can be controlled by adjusting temperature, velocity and direction of air flows in the facility environment, so as to get the optimal distribution of each environmental factor influencing plant growth, such as air, temperature, etc. Embodiment 1 [0029] The present invention provides a plant-caring apparatus and a method thereof to be used in a plant factory for monitoring growth status of plants planted in a cultivation space of the plant factory and performing fine control and fine adjustment on the growth environment. The plant-caring apparatus comprises and the method thereof uses a shuttle cart 1, a care-related collecting unit 2 mounted on the shuttle cart 1, a processing module 3, a monitoring module 4, a data-analyzing unit 5, a control unit 6, and a displacement component 7.

[0030] In a particular mode as depicted in FIG. 1 and FIG. 2, the shuttle cart 1 moves around a cultivation rack 8 in the cultivation space and serves to transfers cultivation plates on which plants are planted to designated locations on the cultivation rack 8. The shuttle cart 1 is further installed with the care-related collecting unit 2 that is configured to recognize growth status of the plants on the cultivation rack 8. The care-related collecting unit 2 can be upheld by the shuttle cart 1 to different positions with respect to the cultivation rack 8, so that the care-related collecting unit 2 is allowed to access the plants at different spatial locations on the cultivation rack 8 and record their growth statuses. The care-related collecting unit 2 then can transmits data it collects about the growth status of the plants to the processing module 3 for comparison, thereby allowing the plants on the cultivation rack 8 that have growth problems to be recognized and marked. The monitoring module 4 collects and monitors environment parameters related to the microenvironments in which leaves of the plants with abnormal growth are according to the marking results provided by the processing module 3. The monitoring module 4 then sends data of the environment parameters it collects to the data-analyzing unit 5. The data-analyzing unit 5 analyzes the data of the environment parameters collected by the monitoring module 4, and according to the analysis results about the microenvironments adjusts the microenvironments of the problem leaves or changes locations of the problem plants on the cultivation rack 8, so that each of the problem leaves or the problem plants can be relocated in a microenvironment facilitating its growth.

[0031] Preferably, the displacement component 7 at its moving end installed with a care-related collecting unit 2 that can capture images of the plants. As shown in FIG. 2, the care-related collecting unit 2 moves with the shuttle cart 1 around the cultivation rack 8 along a predetermined route while capturing images of all the plants on the cultivation rack 8. Preferably, the displacement component 7 can further adjust the relative locations of the care-related collecting unit 2 and the shuttle cart 1, so that when the shuttle cart stops advancing, the displacement component 7 can align the care-related collecting unit 2 with the plant currently to be monitored, thereby allowing the care-related collecting unit 2 to capture images of the plant at different angles. Preferably, when moving with the shuttle cart 1 for the first time, the care-related collecting unit 2 keeps a constant working position with respect to the shuttle cart 1. When receiving data of images of the plants of the same species planted as the same batch on the entire cultivation rack 8 captured by the care-related collecting unit 2 in a given time period, the processing module 3 compares all the images of the plants to each other, and marks the images of the plants that have leaves showing abnormal conditions such as those being wilting, lodging, chlorotic or yellowing.

[0032] Preferably, the shuttle cart 1 takes the care-related collecting unit 2 to follow the same route for the second time. When reaching a place close to a plant corresponding to the marked images, the shuttle cart 1 calls, and the displacement component 7 drives the care-related collecting unit 2 to change its working positions so that the care-related collecting unit 2 can capture images of the plant marked and suspected to be abnormal at multiple angles. These images are to be used for reconfirmation of the abnormal conditions of the plant.

[0033] Preferably, during the second trip the shuttle cart, the displacement component 7 may further change the working position of the monitoring module 4, so that at the same time the care-related collecting unit 2 captures images of the plant at different angles, the monitoring module 4 can move toward leaves of the plant to collect data of microenvironment parameters with respect to the leaves. Preferably, the double heterogeneous verification data used for follow-up confirmation of a suspect plant are data of the images of the marked plant and the microenvironment parameters related to the leaves collected by the care-related collecting unit 2 and the monitoring module 4, respectively. Preferably, the high-definition images of the suspect plant captured at different angles by the care-related collecting unit 2 during its second trip and the data of the microenvironment parameters related to the leaves collected by the monitoring module 4 can be together sent to the data-analyzing unit 5. Preferably, the data-analyzing unit 5 determines whether any of the leaves shown in the images of the plants captured by the care-related collecting unit 2 is wilting, lodging, chlorotic or yellowing by means of image thresholding or contour shape comparison. Preferably, the data-analyzing unit 5 may further compare data of a microenvironment parameter related to the leaves of the plant collected by the monitoring module 4 and a corresponding environment parameter value preset for the cultivation space to find the difference and sends the analysis results to the control center of the plant factory, so that the control center controls the lighting module, watering module, and airing module on the cultivation rack 8 to impact microenvironment around the leaves of the problem plant in terms of lighting intensity, humidity, wind force, and wind direction, and thereby makes the microenvironment favorable to the leaves of the plants for growth and photosynthesis.

Embodiment 2 [0034] The present embodiment provides further improvements on Embodiment 1, and repeated details are omitted from the description thereof [0035] Preferably, the care-related collecting unit 2 is mounted on the shuttle cart 1 through the displacement component 7. The displacement component 7 is configured to be controlled by the control unit 6 to change the working position of the care-related collecting unit 2. After initially marking the images of a suspect plant, the processing module 3 sends data coordinates of the plant on the cultivation rack and corresponding driving commands to the shuttle cart 1 and the control unit 6. The shuttle cart 1, after receiving the coordinate data of the plant from the processing module 3, takes the care-related collecting unit 2 to begin the second trip along the same route, so that the care-related collecting unit 2 performs a new round of normal image collection on non-marked plants and sends data of these images to the processing module 3 for comparison. When the shuttle cart 1 is close to the coordinate location of the suspect plant, it calls, and the control unit 6 controls the displacement component 7 to move along a preset route according to the driving command, so that the care-related collecting unit 2 mounted on the displacement component 7 can capture image of plants suspected with problematic growth status at multiple angles. The displacement component 7 may also drive the monitoring module 4 to change its position, so that the monitoring module 4 can collect data of environment parameters of the microenvironment in a way that it simulates the current growth morphology of the leaves.

[0036] Preferably, the monitoring module 4 is driven by the displacement component 7 under the control of the control unit 6 and moves to different locations on the cultivation rack 8, so that the monitoring module 4 can collect the environment parameters of the microenvironment in the area having a plant with abnormal conditions at a place as close as possible to the actual growth location of the leaves of the plant. The control unit 6 then can, according to the location of the plant marked by the processing module 3 as having abnormal conditions, control the displacement component 7 to drive the monitoring module 4 that simulates the shape of leaves to move toward the leaves growing at the middle section of the plant, so that the monitoring module 4 can collect the environment parameters related to the microenvironment in which the leaves are present in a way that it simulates the morphology of the leaves. Preferably, the monitoring module 4 is driven by the displacement component 7 to perform rotation and shift at various angles with respect to the shuttle cart, so that the monitoring module 4 can change its position around the plant.

[0037] Preferably, the processing module 3 performs recognition and discrimination by 5 determining whether the leaves of the plants in the images are obviously wilting or lodging and comparing the plants in the images with each other in terms of size and color. Preferably, the processing module 3, when comparing the images of the plants collected by the care-related collecting unit 2 at the first time, can determine the order of the images of the plants according to the travel route preplanned for the shuttle cart I, thereby being able to supplement information 10 about the coordinates of suspect plants in the images while making these plants. In this way, the control unit 6 can be informed of the coordinate locations of the suspect plants on the cultivation rack 8 and make the shuttle cart 1 advancing along the preplanned route stop at places corresponding to the suspect plant.

[0038] Preferably, the data-analyzing unit 5 may be preloaded with a sample database that contains parameter values representing average growth status of the plants in a growth cycle. The parameters may, for example, include the average heights of a plant in its different phases and corresponding threshold ranges for the plant in every growth cycle; the aggregate contour area of leaves from overhead projection of a single plant; average diameters of stems of the plant in different phases and threshold ranges thereof; colors and sizes of leaves at the middle section of the plant; whether leaves of the plants are obviously wilting, lodging etc. Preferably, the data-analyzing unit 5 may also be preloaded with a sample database that contains microenvironment parameters related to leaves representing the best growth status of the plant in various phases. Preferably, data in the sample database at least include environment parameters impacting photosynthesis of leaves, such as the temperature, humidity, lighting, water, and CO2, as well as the wind direction and wind force that allow the foregoing environment parameters to happen around the plant in non-uniform distribution.

[0039] Preferably, the control unit 6 is configured to, according to patrol commands regularly given by the control center of the plant factory or driving commands generated by a user initiatively, control the shuttle cart 1 to patrol around the cultivation rack 8 to enable inspection and observation on the plants. When there is not an abnormal plant detected, the control unit 6, at preset cycles, continuously drives the shuttle cart 1 to patrol around the cultivation rack 8 to perform inspection on the plants. When the shuttle cart 1 controlled by the control unit 6 for the first time detects a plant suspected of abnormality, the control unit 6 drives the shuttle cart I immediately to conduct second patrol and inspection. Preferably, the control unit 6, in the second trip of patrol and inspection, may also control the shuttle cart 1 to stop at designated locations along the route, and drives the displacement component 7 to place the care-related collecting unit 2 and the monitoring module 4 into different working positions, thereby obtaining the images of the plants and microenvironment parameters that can be used in follow-up confirmation for the initial results.

[0040] Preferably, when the data-analyzing unit 5 determines that the plant does have growth-related abnormal conditions on the ground that significant differences exist between the images of the plant captured at the second time and the preloaded benchmark data corresponding to the same phase of growth, the data-analyzing unit 5 further determines whether there is any chlorosis, yellow, wilt or lodge caused by abnormalities in the environment where the leaves are present according to the difference between the microenvironment parameters related to the leaves collected by the monitoring module 4 and the benchmark values of the parameters. If the data-analyzing unit 5 after processing the images of the plant captured at the second trip finds that the leaves of the plant has any kind of disease-related abnormal conditions on its leaves such as white deposit, mold, and white spots on the leaves, the data-analyzing unit 5 may further upload the images of the plant to a back-end processing platform on which the control center of the plant factory is located, so that staff up there can perform manual analysis of these conditions for 5 diagnosis or that the processing platform can recognize the conditions shown in the images of the plants against data available through the Internet or available case reports of counterpart plants. This allows further treatment to be conducted according to known solutions to these conditions. For example, the shuttle cart 1 can be driven to move the plant to a different location, or the lighting module, the watering module, and the airing module on the cultivation rack 8 can 10 be controlled to improve the environment around the leaves.

[0041] Preferably, the displacement component 7 comprises a first robotic arm 71 that drives the care-related collecting unit 2 to change its working positions in a certain area and a second robotic arm 72 that pose the leaf-like monitoring module 4 according to the actual working positions and distribution of leaves at the middle section of the stem of the plant for collection of data related to the microenvironment parameters. Preferably, the first robotic arm 71 and the second robotic arm 72 may retract to be received in an accommodating chamber formed in the shuttle cart 1, so that when the shuttle cart 1 performs normal patrol and inspection, the monitoring module 4 remains off. Preferably, the first robotic arm '71 lifts the care-related collecting unit 2 over the shuttle cart, so that when the shuttle cart moves along the preset route, the care-related collecting unit 2 continuously collects images of the plants on the cultivation rack 8 alongside the route. Preferably, the displacement component 7 can fold according to practical monitoring needs for storage. This helps to prevent oxidization, rust, erosion and other degrading reactions of the exposed components in a highly humid or microbe-affected environment, thereby maximizing service life and durability of components, especially of the robotic arms.

Embodiment 3 [0042] Preferably, the processing module 3 and the data-analyzing unit 5 process and analyze the images of the plants may further include comparing leaves at upper, middle, and lower sections of the stems of the plants in the images in terms of color and contour dimension. When the comparison suggests that a certain plant or a few plants has its or their contours significantly smaller than the other plants in the images, further analysis may be performed to ascertain the location of this plant or distribution of these plants, thereby determining whether the abnormal conditions are caused by fungus that survived the incomplete cleaning of a cultivation plate before cultivation, or the growth retardation is caused by an improper formula of the nutrient solution supplied to this area of the cultivation rack 8 or local deficiency of some certain nutrients, or the growth restriction is caused by a local electric or lighting problem that has prevented the lighting module from providing lighting effectively to the plants in the affected area. Preferably, the lighting problem may refer to a broken light, an unsatisfying lighting color that is not particularly favorable to growth of the plants, or over-bright/dull lighting. Preferably, other factors that cause differences between plants in terms of growth may include an improper location of an air vent or air outlet on the cultivation rack 8 and a wrong level at which the cultivation plate is placed on the cultivation rack 8.

[0043] A suitable environment is essential for an organism to live and develop. For example, a plant needs the appropriate temperature, humidity, lighting, water, and CO2 to perform photosynthesis or its living and development are impossible. Herein, a microenvironment, also known as a microdomain, is where a leaf is present. Optimization of a microdomain around a leaf only requires maintaining environmental factors in a very small area at the surface of a leaf within an optimal range, without controlling the temperature and humidity of the entire cultivation space or the entire cultivation rack 8. This is because no matter how the climate changes, what really affects development of a plant is the temperature at the surface of each of its leaves and the air humidity in a 0.5 cm distance from the surface of each of its leaves. As 5 proven in experiments, the temperature and humidity in the microenvironment of leaves may be very different from those in a seedbed space or from the atmospheric temperature and humidity. Superficially, in hot seasons, when the air humidity of at 0.5 cm reaches 90%, the air humidity in a 1.6 cm distance from the leaf surface is merely 40%. Fortunately, to keep the air humidity at 0.5cm above 90% is not difficult and may require just a few fixed atomization nozzles, without 10 high consumption of electricity and water.

[0044] Preferably, the monitoring module 4 simulates leaves by reproducing the stoma structure of the plant and its requirements for water metabolism and photosynthesis temperature in an isolated state. The dummy is in the form of artificial leaves made of high-density circuits, special materials, and sensors. It can sense various factors in the microenvironment around leaves of a plant, such as the temperature, the evapotranspiration coefficient, water film distribution on a leaf, substance moisture, air humidity, substance humidity, ion concentrations of minerals (EC values), environmental lighting, and other environment parameters. The data-analyzing unit 5 processes data of these microenvironment parameters fed back by the artificial leaves, and uploads the analysis results about plants with growth abnormality to the control center. The control center then according to the analysis results controls implementation mechanisms to turn on/off devices peripheral to the plant growth environment, thereby modulating the external environment.

[0045] The control unit 6 controls the monitoring module 4 and the second robotic arm 72 in two-level hierarchy. The first level is a direct control level, which is in the form of a sub-controller based on a single-chip microcomputer, e.g., AT89C51. It controls the monitoring module 4 that has a leaf-simulating structure to monitor environmental factors such as the temperature, humidity, lighting and CO2. Preferably, the control unit 6 is further configured to perform criticality control on greenhouse devices in a real-time manner according to the monitoring results from the monitoring module 4. Herein, the term "criticality control" refers to an operation wherein if the measurement of a certain microenvironment parameter related to the growth environment of a plant has a value between upper and lower limits set for the very microenvironment parameter, no control actions will be taken, and if the value is below the lower limit or above the upper limit, relevant control devices will be activated. Such an operation is favorable to energy-saving purposes. Criticality control is known to be an economic yet productive means of control. Herein, the implementation mechanisms may include water pumps, ground-heating wires, red-light plant growth lamps, etc. Preferably, the second level is a process management level, which is realized using a plant growth controller, such as one built from a W78E58B single-chip microcomputer. It serves to manage the environment parameters and configure control parameters. The system can be easily modified and expanded as needed. The plant growth controller is prepared with an AID conversion circuit for processing data fed in by sensors. It may be designed as a separate control system for handling growth-related tasks in a designated area.

[0046] Preferably, the monitoring module 4 is built to simulate leaves of a plant so it is able to mimic how the surface of a leaf is exposed in the environment during growth of the plant. Preferably, the monitoring module 4 integrates the sensors logging the environment parameters, i.e., the temperature, humidity, lighting, nutrition, CO2, and dissolved oxygen, into a leaf-shaped structure. Preferably, the leaf-simulating monitoring module 4 simulates leaves of the plant in terms of stoma structure, water metabolism, photosynthesis, and temperature requirement. It is bult from high-density circuits, special materials, and sensors, and is in the form of sensing-enabled artificial leaves, so that it can measure various factors in the microenvironment around leaves of the plant, including temperature, the evapotranspiration coefficient, water film distribution on a leaf, substance moisture, air humidity, substance humidity, ion concentrations of minerals (EC values), environmental lighting, and other environment parameters. The control center of the plant factory introduces the parameter measurement obtained by the dummy leaves into a rapid propagation system for computing, and instructs automated implementation mechanisms to turn on/off peripheral devices accordingly, thereby modulating the external environment. By acting as a sensing organ that monitors variations of the environmental factors (microenvironment parameters) during growth of the plant, the monitoring module 4 can precisely acquire data related to parameters that have impacts on interaction between leaves and the environment and achieve systematical data collection. This allows the monitoring module 4 to accurately simulate the environment around the leaves. Particularly, when significant growth difference between plants in the same area or between parts of the same plant is detected from the images of plants acquired by the care-related collecting unit 2 mounted on the shuttle cart 1, the monitoring module 4 can be upheld at different leaf positions to acquire information about the environment parameters regarding leaves showing different growth status while simulating the environment around the leaves, thereby enabling the data-analyzing unit 5 to perform analysis on the parameters that are potentially responsible for growth abnormality.

[0047] Preferably, the monitoring module 4 further comprises an infrared CO2 analyzing unit that measures CO2 generated by the plants during respiration and thereby determines the photosynthetic rate of the plants. Specifically, when the shuttle cart 1 performs patrol and inspection at night, the care-related collecting unit 2 may be prevented from capturing clear images due to darkness, whereas the infrared CO2 analyzing unit in the monitoring module 4 can determine whether the plants in a given area are breathing normally by measuring the CO2 level in the air, so as to recognize the plants in areas where the CO2 level is low and conduct further measurement and sampling. Preferably, with effective analysis of the CO2 level enabled by the infrared means, growth of the plants can also be well monitored at night. This perfectly complements the care-related collecting unit 2 that is incapable of capturing clear images at night and enables more accurate analysis.

[0048] Preferably, the monitoring module 4 that is leaf-shaped and positioned like leaves of the plant can further detect environmental variations related to leaves in a real-time manner. For example, after spray irrigation has been conducted in the cultivation space for cooling the plants, the monitoring module 4 can determine that proper environmental modulation is done when it detects that the air temperature around the surface of leaves of the plants to be cooled has lowered to a certain extent. This is superior to the prior-art solution that requires the environment around the whole plant or in the entire plant factory to be cooled, thereby significantly improving modulation efficiency and reducing energy consumption in the plant factory for growing the plants. Thereby, the control center of the plant factory or the control unit 6 can perform environmental modulation in a relatively small scale according to the parameter data from the monitoring module 4 and change the growth environment of the plants corresponding to the parameter data collected by the monitoring module 4.

Embodiment 4 [0049] The present invention further provides a plant-caring apparatus for a plant factory.

[0050] The disclosed apparatus comprises a shuttle cart 1, a care-related collecting unit 2, a processing module 3, a data-analyzing unit 5, a cultivation rack 8, and a growth-facilitating device 9.

[0051] In the following embodiment, some terms involved are specified as follows.

[0052] Herein, the shuttle cart 1 and a patrol cart 310, a shuttle vehicle as well as a plant-inspecting patrol robot as used in the following embodiment all refer to the same device; [0053] Herein, the care-related collecting unit 2 and a visual module as well as an image-capturing unit as used in the following embodiment all refer to the same module; [0054] Herein, the processing module 3 and a processor in a plant growth monitoring device 300 15 as used in the following embodiment both refer to the same module; [0055] Herein, the data-analyzing unit 5 and a data processing center 320 as used in the following embodiment both refer to the same module; [0056] Herein, the cultivation shelf 8, a growing rack, and a vertical cultivation shelf all refer to the same structure; and [0057] Herein, the growth-facilitating device 9 and a nutrient solution unit as used in the following embodiment both refer to the same module.

[0058] Herein, the plant growth monitoring device 300 comprises a patrol cart 310 and a data processing center 320. Therefore, the plant growth monitoring device 300 is a shuttle cart 1 equipped with the data-analyzing unit 5, and the image-capturing operation to be performed by the patrol cart 310 is accomplished when the care-related collecting unit 2 moves with the shuttle cart 1 to a designated location and takes pictures of plants.

[0059] As shown in FIG. 2, the cultivation rack 8 is constructed from multiple layers of floating shelves. For each shelf, at least one cultivation area and a section of a patrol track are provided. The cultivation area is where plants are cultivated, and the patrol track is routed around the cultivation area.

[0060] The patrol track is arranged on every shelf of the cultivation rack 8, so that the patrol cart 310 can patrol throughout every shelf of the cultivation rack 8 by moving along the track. Preferably, the patrol track zigzags around the cultivation rack 8 and goes through all the shelves continuously, so that the patrol cart 310 can be guided to visit every shelf of the cultivation rack 8 without interruption. The patrol cart 310 thus supports dynamic, real-time monitoring of the plans through infrared inspection and image comparison. Preferably, the patrol cart is equipped with infrared cameras, monochrome video cameras, chromatic video cameras, and a processor.

[0061] The patrol cart 310 has a visual module and a processor. When the patrol cart 310 moves along the patrol track, the visual module captures images of the plants in the cultivation areas. These images then provide a basis for assessment of plant growth. Preferably, the visual module captures both monochromatic and chromatic images. The processor on the patrol cart 310 is capable of extracting and comparing data of monochromatic images at a relatively simple level.

When there is any abnormal condition detected, it sends a command to take chromatic pictures.

Preferably, the patrol cart 310 is configured into a gantry-like structure so that it can span the nutrient solution tank. Its two sides are each provided with a lifting unit for lifting cultivation plates. The gantry-like structure of the patrol cart 310 is slidably connected to the track through lateral supports, and it comprises a bar transversely connecting the two vertical lateral supports.

Such a structure allows the track and the nutrient solution tank to be arranged in parallel and the clearance to the ground defined by the supports provides a space for receiving the nutrient solution tank when the patrol cart 310 passes the nutrient solution tank from above. Moreover, the supports at the two laterals can be adjusted in height by the data processing center 320 through retraction or extension. When the patrol cart 3 1 0 approaches the cultivation rack 8, the supports can extend to lift the bar to the altitude corresponding to a given shelf of the cultivation rack 8, so that the visual module installed on the bar can scan plants on the shelf of the cultivation plant for monitoring.

[0062] The data processing center 320 is in data communication with the cultivation rack 8 and the patrol cart 310 wired or wirelessly. The data processing center 320, in response to a command it receives or data fed back to it, makes commands. The cultivation rack 8 then follows the command from the data processing center 320 to change setting of lighting or other facilities on the cultivation rack 8. The visual module of the patrol cart 310 takes pictures of plants on the cultivation rack 8 it passes by. The processor on the patrol cart 310 or the data processing center 320 then uses these pictures for characteristic extraction. By comparing the characteristics of the currently monitored plants as shown in the captured images with those of normal plants, growth abnormality of the currently monitored plants can be found and then related diagnosis can be performed. Preferably, the data processing center 320 performs coupled control on multiple environmental factors through man-machine interface, thereby achieving better control efficiency and precision.

[0063] According to a preferred mode of the present invention, the patrol cart 310 is configured to take monochromatic pictures, gray-scale pictures, and chromatic or true-color pictures according to data loads and requirements for extraction of trait characteristics. Herein, abnormal conditions, or abnormality, of plants mainly refer to illness preliminarily caused by diseases or pests. A plant affected by diseases and/or pests is likely to show related signs at its leaves or stem, such as malformation and/or discoloration of tissues. In other words, the shape and color of a plant tissue provide can be used as indicators that a plant is with any abnormality or disease or not. To this end, monochromatic pictures can be very useful. For example, an ill plant may have shriveled leaves. The patrol cart 310 takes monochromatic pictures of the leaves of the plant, and extract characteristics on coordinate points along the contours of the leaves of the plant. Decrease in leaf stretch and in area as proven by difference of leaf contours can indicate that the subject plant is with abnormality. Preferably, Canny may be used for edge detection, and then a minimum bounding box or a minimal radius bounding semi-circle may be used for final determination.

[0064] Further, when any abnormality is confirmed in a monochromatic image of a plant, the patrol cart 310 selectively tales gray-scale, chromatic, or true-color pictures of plants in the corresponding area. As shown in Table 1, to determine whether there is any plant with abnormality/illness, selection of the suitable image type may be made according to the abnormality/illness detected. Herein, a monochromatic picture is preferably of 1600 * 900 * 2 bits.

[0065] Table 1

Influencing Tissue Pest Pest Damage Pest Damage Pest Factor Morphology Damage Morphology Injury Color Damage Location Injury Size Image Type Monochromatic Gray-scale Gray-scale Chromatic /True Color Gray-scale [0066] Gray-scale pictures are suitable for extraction of plant characteristics related to pest damage, such as pest morphology, injury distribution, and affected area sizes. For distribution of pest damage on some certain plants, chromatic or true color pictures may be further taken.

Preferably, the patrol cart 310 move at a high speed can use its built-in processor to perform characteristic comparison based on monochromatic pictures. Advantageously, the patrol cart 310 timely detects pest damage on site can immediately suspend its trip to take more gray-scale pictures of the plants having pest damage successively at this spot. As a more serious consideration, the patrol cart 310 encountering plants with pest damage is likely to carry pest eggs. In this case, if the contaminated patrol cart 310 continues its trip, it may propagate the pests to other plant areas, making the pest problem spread. Therefore, when a patrol cart 310 finds any plant suspected of pest problems, it stays in the current area util proper protective measures such as bagging or measures have been done.

[0067] After the pest problem affecting the plant is preliminarily ascertained according to the morphology, color and locations of the leaf spots, gray-scale pictures of the front and back of the plant may be further taken for a more comprehensive analysis of distribution and area of affected tissues. The analysis result, combined with color shade of the leaf spots, can indicate the severity of pest damage of the plants and its influence of the yield of the plants. For example, the tomato fruitworm mainly affects fruits of the crop, and seriously reduces the yield the yield. Tomato fruitworm larvae mainly consume buds, flowers, and fruits, and they also damage stems, leaves, and shoots. When a bud is affected, its bracts loose and turn to yellowish green. The bud falls off after 2 to 3 days. The young fruits are often eaten out from inside or fall off due to rottenness. While adult fruits are often partially eaten, the wormholes made at the stalks can provide access to the interior for external water and bacteria, and the fruits eventually rot and fall off, leading to decrease in production. In its proliferative and growing phases of tomato, tomato fruitworm individuals are found in the buds or at the surface of the fruit. The patrol cart 310 make take gray-scale pictures of the plants with pest damage, and use gray-scale differences between edges of buds, fruits and worms in the pictures to discriminate them, thereby determining the distribution, scope, and morphology of the worms on the plants.

[0068] If there is problem to recognize the species of the pests found at the surface of the plants, chromatic or true-color pictures may be taken to provide clear images of the pests as the basis for the data processing center to make pest recognition.

[0069] With information of the pest species, the pest damage distribution, the pest damage area, and the current state of the pests, health of the plants can be evaluated. The plants affected by the pests have to be treated with pesticide, and information of their current damaged states can be recorded as benchmark data for assessment of subsequent treatment so that the data processing center 320 can reasonably select the suitable pesticide, its dosage, and its administration accordingly.

[0070] The data processing center 320 can evaluate the abnormal conditions of a plant by performing characteristic extraction on pictures of the plant. Herein, different types of pictures form a hierarchy for determination of pest damage of the plant. The plant factory uses and interpret information obtained at different levels in the hierarchy differently. For example, monochromatic pictures can be simply processed by the processor built in the patrol cart 310, without involving resource-consuming data transmission and the high-precision, high-speed computing power of a processing center. This allows low-cost and fast data processing, so that the patrol cart 310 can continuously scan plants to find abnormality as it moves along the track in the cultivation area. On the other hand, there are plants that have their leaves or stems as their produce. For these plants, once their leaves or stems are affected by pests, their values are in jeopardy. For example, the species Pieris rapae can damage plants at leaves and make the affected plants lose their value as herbal medicine. When such pest damage is found, the pressing matter of the moment is to isolate the affected plants from healthy plants, but not to consider whether gray-scale pictures or chromatic pictures should be taken.

[0071] The data processing center 320 and the processor process the pictures taken and physiological data regarding the current status of the plants can be obtained. The physiological data dynamically reflect the environment in the plant factory. Most known approaches in the art give more emphasis to factor-based environmental adjustment for the plant factory itself, and fail to see that the objective of environmental modulation is to best serve plants and facilitate their growth. For example, China Patent Publication No. CN112470790A provides a plant growth environment monitoring and adjusting apparatus and its method. The apparatus comprises a planting greenhouse, partitions, detection units, adjusting units, and control units. The planting greenhouse is divided into a plurality of planting rooms with the partitions. Each of the planting rooms is equipped with a detection unit. The detection unit measures planting environment factors in the planting rooms and sends the measurements to the control unit. The control unit then records growth environment data according to plant growth status and determines a combination of growth environment factors that is most favorable to plant growth through matching, and controls the adjusting unit to change the growth environment in the planting room according to the combination of determined factors. The plural planting rooms can be monitored and adjusted at the same time. The known apparatus measures growth status of plants according to contents of gases in the environment, and modulates temperature and lighting accordingly. However, what can represent the growth status of a plant most directly is the condition of the plant itself, not contents of gases in an environment. In other words, contents of gases in an environment represent merely one of the factors that have impacts on growth status of plants. Therefore, the present invention uses the patrol cart 310 to scan images of plants so as to enable intuitive determination of health of the plants in terms of nutrition and reproduction.

[0072] The data processing center 320 processes data with at least three priority levels, including 35 Level 1 associated with monochromatic pictures, Level 2 associated with gray-scale pictures, and Level 3 associated with chromatic pictures.

[0073] The patrol cart 310 captures images of plants by taking monochromatic pictures of Level 1 priority. Therein, the patrol cart 310, while moving along the track at a high speed, uses cameras to generally scan plants coming into its field of view. The data processing center 320 or the processor of the patrol cart 310 then performs characteristic extraction and comparison on the pictures taken during the foregoing scanning to compare morphology contours of stems and/or leaves, thereby determining whether there is any plant with abnormality. Therein, a plant with abnormality may be a plant suffering from nutrition imbalance or lighting imbalance and pest damage. When a plant with abnormality is found, the patrol cart 310, in response to the command issued by the data processing center 320 about taking gray-scale pictures of Level 2 priority, captures images of the plant with abnormality by enhancing color richness of the images captured. The gray-scale pictures corresponding to Level 2 priority can be used as a material from which characteristics of disease-caused spots or pest damage, such as their morphology, locations and sizes on plant stems and/or leaves, can be extracted.

[0074] If the analysis results, namely morphology of the disease-caused spots or pest damage, are not sufficient for the data processing center 320 to recognize the source of the disease-caused spots or pest damage with the related data in its database, the data processing center 320 commands the patrol cart 310 to take pictures of Level 3 priority. The Level 3 pictures provide information if existence and distribution of foreign objects.

[0075] According to a preferred mode of the present invention, the patrol cart 310 preferentially uses monochrome video cameras to scan and record plants on the cultivation racks as it moving at a high speed. A monochromatic picture requires less resources and capacity to process as compared to a gray-scale picture, a chromatic picture, and a true-color picture. As the preliminary objective of processing is to compare plant contours and morphology, a monochromatic picture is completely competent. The monochromatic pictures of plants taken by the patrol cart 310 in a scanning-like manner are relatively small in volume, and thus the built-in processor of the patrol cart 310 can process these pictures to extract morphologic characteristics, such as contours, of leaves and stems of the plants therefrom and determine whether the plants are with abnormality by comparing the morphologic characteristics to those of healthy plants.

[0076] After the area in which one or more affected plants are found has been determined, the 20 next step is to ascertain the severity of the related disease and groups of affected areas. The group information is helpful to selection of treatment schemes, and the severity information can be used as benchmark data for subsequent treatment.

[0077] Determination of spot groups can be accomplished with reference to information about colors, locations, and shapes of disease-caused spots on the affected plants. Therein, the locations and morphology of the disease-caused spots can be determined using the gray-scale pictures of the plants taken by the patrol cart 310, and the colors of the spots are recognized in chromatic or true color pictures of the plants taken by the patrol cart 310. It is known that, from the data-processing perspective, a gray-scale picture is smaller in volume than a chromatic or true color picture, and a monochromatic picture is smaller in volume than a gray-scale picture. With this fact, it is preferred that the type of the highest priority level is preferentially adopted when pictures are taken as long as the purpose of morphological comparison can be satisfied. Specifically, a monochromatic picture is prior to a gray-scale picture that is prior to a chromatic picture that is prior to a true-color picture.

[0078] Preferably, the patrol cart 310 comprises at least a first track-guided cart and a second track-guided cart. The first track-guided cart has a visual module equipped with a monochromatic image-capturing unit for taking pictures of Level 1 priority. The first track-guided cart, when moving along the track, can at least use cameras that can recognize growth status of plants to preliminarily determine the growth status of all plants planted in all cultivation plates on all cultivation shelves of all cultivation racks.

[0079] The visual module of the second track-guided cart is provided with an image-capturing unit that captures at least gray-scale images. The first track-guided cart and the second track-guided cart can both follow the track to move on and around the cultivation racks 8, so as to ensure that the second track-guided cart can reach the place located by the first track-guided cart, thereby precisely locating all affected plants. When plants in a certain cultivation area are determined as having abnormal growth status according to information from monochromatic pictures, the image-capturing unit on the second track-guided cart starts and take pictures of plants in the cultivation area that is problematic as suggested by the monochromatic pictures, so as to further recognize the related diseases and/or pests and their severity. Preferably, the data sent to the data processing center 320 by the first track-guided cart include abnormal conditions of the plants, locations of the abnormal plants, and locations of the first track-guided cart. The data processing center 320, with knowledge of the locations of the abnormal plants and the locations of the first track-guided cart, can send positioning commands to designate the travel route and stops of the second track-guided cart, and it also assists the second track-guided cart in determining the scanning directions. For example, when the first track-guided cart arrives at the first location in the first cultivation area, and the image-capturing unit detects abnormal plants at locations corresponding to Row 5-Column 4, Row 5-Column 5, and Row 5-Column 6 in the first cultivation area, the second track-guided cart can move to the first location according to information from the data processing center 320, and identify the plants at locations Row 5-Column 4, Row 5-Column 5, and Row 5-Column 6 in the first cultivation area to perform its Level 2 image capturing operation.

[0080] The image-capturing units for recognizing growth status of plants are each in data communication with the control unit of the smart plant factory through a first communication unit built in the respective track-guided carts. Preferably, the first track-guided cart can move along the track and perform preliminary recognition of the growth status of all plants planted in all cultivation plates on all cultivation shelves of all cultivation racks regularly or irregularly.

[0081] The first track-guided cart can use its built-in processor to process the monochromatic pictures it captures and detect abnormality without sending the monochromatic pictures to the control unit or another module for recognition and analysis. Since the first track-guided cart processes data locally, efficient data processing and timely feedback can be achieved while the first track-guided cart keeps its fast advancement, thereby allowing the second track-guided can to conduct further inspection efficiently.

[0082] Instead of the priority setup as described above, if the images of the plants in cultivation plates are always captured as chromatic images/video clips that are huge in size, the resulting high memory requirements at the track-guided carts and heavy data-processing loads born by the control unit can eventually lead to inefficient image-based recognition of disease damage and/or pest damage. Consequently, the travelling speed of the carts along the track and the patrol capacity of the system are disadvantageously limited, making the solution inapplicable to large-scale plant factories.

[0083] By using monochrome video cameras to take pictures/video clips or using cameras to take monochromatic pictures/video clips, and making these cameras themselves capable of recognizing growth status of plants, the present invention allows efficient analysis am preliminary determination of disease damage and/or pest damage of plants to be conducted locally. This eliminates the need of transmitting the captured monochromatic pictures to the control unit for processing, and significantly improves overall efficiency of recognition. Further, as the efficiency of preliminary recognition is improved, the travelling speed of the first track-guided cart can be increased significantly. As a result, the plant-monitoring efficiently and capacity of the entire plant factory can be improved, while good accuracy of determining whether plants have abnormal growth status can be assured, making the present invention suitable for large-scale plant factories.

[0084] The second track-guided cart may further be configured to investigate causes of detected abnormal conditions. Particularly, it captures images in a streamlined process, so as to enhance 50 image-capturing efficiency and image-processing efficiency by preventing unnecessary switching operations between different types of pictures or cameras.

[0085] The disclosed system further comprises a temperature-controlling device. The temperature-controlling device is also controlled by the data processing center 320. A plant affected by desiccation or high heat can have its leaves shriveled or turning yellow. When the first track-guided cart detects a plant with abnormality (e.g., shriveled leaves) in its inspection associated with Level 1 priority, the second track-guided cart follows the process up with inspection associated with Level 2 priority. If no foreign objects are detected, the data processing center 320 reviews settings of lighting, temperature, and/or nutrition for the plant, and further informs the data processing center 320 of its finding. The data processing center 320 further checks humidity, temperature, lighting, and nutrition supply for the plant in the growth environment, and controls a growth-facilitating device 9 to modulate relevant environmental factors, thereby ensuring growth abnormality seen in the plant is not a result of any improper environmental factor.

[0086] In a preferred mode of the present invention, the data processing center, according to results of the foregoing inspection and check, controls photosynthesis nutriment nozzles and/or rhizosphere nutriment nozzles to water the area in which the abnormal plant exists by spraying water or nutrient solution to improve the growth status of the plant.

[0087] The cultivation racks 8 may be arranged in a sowing/seedling zone, a planting/thinning zone, and a growing zone corresponding to different growth phases of plants. In the sowing/seedling zone, the cultivation racks 8 are provided with sponge pads and cultivation plates. The seedling sponge pads are for vegetable seeds to be sowed thereon. Specifically, the sponge pads having seeds sowed therein are put in cultivation baskets and arranged orderly on cultivation plates. The cultivation plates are then placed into seedling racks to grow seedlings. Lighting on the seedling racks remain off until the seeds germinate, which typically takes 2 days. After germination, the lamps are activated to illuminate and extinguish as programmed through a time controller. Preferably, each sponge pad is of 25mm x 25mm X 25mm. For different zones, independent air-conditioning systems are used. Preferably, the temperature is 20 °C for the light period and 18 °C for the dark period. The shelves on the cultivation racks in the sowing/seedling zone are each equipped with a nutrient solution circulating system and a lighting system. Preferably, fluorescent lamps or white LEDs are used as light sources, with luminous intensity of 10 p.m ol -m-2.s-1.

[0088] In the planting/thinning zone, the cultivation racks 8 are provide with seedling racks. Seeds are cultivated on the seedling racks for about 15 days. When the seedlings reach a proper size, they are transferred to a planting zone where they are planted. The planting operation is about increasing intervals between seedlings by transferring them from a seedling plate having a high-density setup to a seedling plate having a low-density setup. The emptied seedling plate can be cleaned and stored in a buffering zone until its next use. Seedlings planted on the small-seedling plates are moved to the cultivation racks in the growing zone on for further cultivation for 15 more days. Afterward, they are taken out from the cultivation racks and sent to a thinning station for thinning. Herein, the thinning operation is about taking small seedlings out of a small-seedling plate and moving them to a big-seedling plate that features an even lower density setup that provides more growth space for the seedlings. The emptied small-seedling plates are cleaned and stored in the buffering zone until the next planting operation. The small seedlings moved to the big-seedling plates are then transferred to the cultivation racks in the growing zone of for further cultivation.

[0089] In the growing zone, the cultivation racks 8 are placed in a triple-height artificial-light cultivation room. Preferably, the artificial-light cultivation room has a net height of 93 m.

[0090] The present invention further involves use of loading/lifting machines, flatbed laser-guided AGVs, planting/thinning robotic manipulators, and high-speed transfer vehicles. The loading/lifting machines take out cultivation plates from the cultivation racks and transfer them to high-speed transfer vehicles. The handling mechanism used may be telescopic forks operatable in both directions. Each of the forks has a water sump formed at its middle part so that the nutrient solution dropping from roots of plants in the currently handled cultivation plate can be collected in the water sump. The forks can move horizontally over the loading platform of a lifting machine to handle two rows of cultivation plates placed on the cultivation racks of the same group. The AGV uses two-wheel differential drive and is powered by a lithium-ion battery. Its features include laser positioning, trackless walking, and automatic lifting. With differential spin ability, it is particularly resistant to uneven and slippery road surfaces and human pushes. When encountering obstacles, it may stop or bypass obstacles according to practical needs and the command from the central management system. The high-speed transfer cart is positioned at the front end of the warehouse and consists of a base and a sliding table that is equipped with a conveyor for conveying cultivation plates. The high-speed transfer cart can be connected with a plurality of devices in series, so as to continuously transfer the cultivation plates. The planting/thinning robotic manipulator is constructed from truss mechanisms to perform planting and thinning operations automatically. The robotic manipulator has its front end provided with clamping jaws for clamping cultivation baskets, so as to facilitate planting and thinning operations. Preferably, the robotic manipulator may be installed on the patrol cart 310 so that it operates while moving with the patrol cart 310 to traverse the cultivation rack or follow the track.

[0091] According to a preferred mode of the present invention, the plant production system to be used in a plant factory operates through the following production process: manually assembling seedling plates, cultivation baskets, and cultivation sponges in a sowing room at the second floor, sowing with a sowing machine, transferring the seedling plates to seedling racks, cultivating the seedlings on the seedling racks for about 15 days, manually transferring the seedling plates to a transport pipe in front of the lifting machine at the second floor with a trolley, sending the seedling plates to the planting zone at the first floor through a lifting machine, sending the small-seedling plates for planting from the small-seedling plate buffering zone to the planting workstation using AGVs, sending the small-seedling plates after the planting operation to designated locations on the cultivation racks using the high-speed transfer vehicles, the lifting machine, and the shuttle carts, sending the emptied seedling plates after the planting operation with AGVs to the cleaning room at the first floor for cleaning, sending the cleaned, stacked seedling plates with AGVs to the entrance of the lifting machine at the first floor, and sending the seedling plates to the sowing room at the second floor through the lifting machine for further sowing use.

[0092] The small seedlings are cultivated in the cultivation area for about 15 days and then the small-seedling plates are sent to the thinning zone by the shuttle carts, the lifting machine, and the high-speed transfer vehicles for the seedlings to receive thinning. Idle big-seedling plates to be used in the thinning operation are transferred by AGVs from the idle big-seedling plate buffering zone to the thinning zone. The thinning operation is automated, and big-seedling plates after the thinning operation are sent by the high-speed transfer vehicles, the lifting machine, and the shuttle carts to the cultivation area for further cultivation.

[0093] In the present invention, the industrial plant production system further comprises a harvesting and packaging zone and a plate cleaning and storing zone.

[0094] The harvesting and packaging zone is where mature plants get harvested. The big seedlings are cultivated in the cultivation area for about 15 days, and the small-seedling plates are sent by the shuttle carts, the lifting machine, and the high-speed transfer vehicles to the harvesting zone where the harvesting operation is conducted. Before harvest, the part of the vegetables exposed below the cultivation baskets, i.e., the roots, is cut off The removed roots are collected in a root collector. Then the big-seedling plates are handled by the harvesting robotic manipulator that performs harvest and moves the produce to the conveying machine. The now emptied, idle big-seedling plates are sent by the high-speed transfer vehicles to the conveying line at the north side and transported by the AGVs to the cleaning room for cleaning. The cleaned plates are transported by the AGVs to the idle big-seedling plate buffering zone. The harvested vegetables are first manually sieved, and only qualified vegetables are sent to the packaging machine where they are packaged, weighted, and labelled. They subsequently are stacked into crates by the robotic manipulator. The vegetables in the crates are sent to a pre-cooling room for pre-cooling and then sent by the AGVs to the packaging room to be prepared for shipment.

[0095] The plate cleaning and storing zone is the place that the reusable cultivation plates after use are cleaned and prepared for reuse. The emptied cultivation plates after their use in a previous procedure are sent by the AGVs to the cleaning room. Here, the plates are cleaned using high-pressure cleaners and dried by high-pressure air. Then they are stacked by the plate-stacking machine and sent by the AGVs to the idle plate buffering zone for later use.

Embodiment 5 [0096] As shown in FIG. 3, the disclosed plant-caring apparatus for a plant factory may further comprise a process unit 10 and a light supply unit 11. The light supply unit 11 is configured to illuminate plants planted on the vertical cultivation rack. The nutrient solution unit is configured to feed the plants with nutrient solution. The process unit 10 is configured to at least receive and/or transport cultivation plates that carry the plants and are to be placed on the vertical cultivation rack. The process unit 10 also serves to further handle and process the plants planted on the cultivation plates. The control unit 6 is configured to at least provide lighting and nutrient solution to the plants through the light supply unit and the nutrient solution unit, respectively, according to the growth needs of the plants, and control the process unit to further handle and process the plants.

[0097] The process unit is configured to at least receive and/or transport the cultivation plate transferred by a transferring mechanism unit, and further handle and process the plants planted on the cultivation plates. The process unit at least comprises plural standalone conveying machines for conveying the cultivation plates. The process unit further comprises high-speed transfer vehicles, planting robotic manipulators, thinning robotic manipulators, plate-stacking machines, root-removing machines, harvesting robotic manipulator, packaging machines, weighting/labeling machines, and parallel-connected robots. Therein, the high-speed transfer vehicles take over the cultivation plates from the loading/lifting machines and transport them to subsequent procedures handled by the process unit.

[0098] Preferably, quality-assurance cameras are used to capture images/video clips of the plants collected by the harvesting robotic manipulators and send the images/video clips to the control unit for determining whether the plants are of good quality. If the control unit determines that the 50 plants have acceptable quality, it allows the plants to enter the subsequent procedures. Otherwise, the control unit controls the harvesting robotic manipulators to transfer the weeded plants to waste bins. Preferably, the control unit can acquire quality check information of the plants harvested by harvesting robotic manipulators from the quality-assurance cameras in a realtime/non-real-time manner. More preferably, the quality check information is at least about the variety of the plants, the historical growth data of the plants, and quality assessment of the plants. Preferably, the historical growth data of the plants may include but are not limited to data of light formulas, nutrient solution, temperature and humidity, carbon dioxide supply the plants have experienced in different phase of the plants from sowing to maturing. Preferably, the quality information of the plants may include but is not limited to: quality ranks of the plants, plant heights, average leaf dimensions, and fruit dimensions and weight. Preferably, the historical growth data of the plants may be acquired from the control unit. With the configuration described previously, the control unit can get hold of quality information regarding plants of each and every batch, and conduct backtracking with the quality check information (particularly the growth historical data of the plants) so as to conclude a more reasonable planting scheme of the variety of plants. For example, adjustment may be made to light formulas, carbon dioxide supply, temperature and humidity, nutrient solution supply for the plants in order to reduce defective produce from the plant factory and thereby improve overall quality of the product (e.g., vegetables) of the plant factory. For example, vegetables of the same variety (such as tomatoes) planted on cultivation plates as the same batch may be not of uniform quality, such as fruits of different sizes and fruits with small diameters in average. In such a case, the control unit traces back the historical data of this batch of tomatoes to figure out whether there is any opportunity to optimize future planting of this variety of plants. For example, the control unit may perform statistic and comparative works may be performed on light formula data, nutrient solution data, temperature and humidity data, carbon dioxide supply data for the variety of plants related to multiple batches so as to conclude a planting scheme more favorable to the variety of plants, thereby allowing the plant factory to produce plants of this variety with improved quality.

[0099] It is to be noted that the particular embodiments described previously are exemplary. People skilled in the art, with inspiration from the disclosure of the present disclosure, would be able to devise various solutions, and all these solutions shall be regarded as a part of the disclosure and protected by the present disclosure. Further, people skilled in the art would appreciate that the descriptions and accompanying drawings provided herein are illustrative and form no limitation to any of the appended claims. The scope of the present disclosure is defined by the appended claims and equivalents thereof Throughout the disclosure, any feature following the term "preferably" is optional but not necessary, and the applicant of the present application reserves the rights to withdraw or delete any of the preferred features any time.

Claims

What is claimed is: I. A plant-caring apparatus, which is configured to be used in a plant factory, the plant-caring apparatus being characterized in comprising: a care-related collecting unit (2), being configured to monitor growth status of plants through collecting images of the plants by means of infrared imaging and/or visual imaging; and a processing module (3), being configured to compare the images of the plants with each other so as to recognize and mark any said plant that exhibits an abnormal condition in the growth status, and control the care-related collecting unit (2) and the monitoring module (4) to collect double heterogeneous verification data for a follow-up confirmation operation performed on the marked plant, wherein the care-related collecting unit (2) changes a working position thereof so as to shift itself among a plurality of observation angles around the marked plant while collecting the data for the follow-up confirmation operation; and the monitoring module (4) changes a working position thereof in a way that it simulates working states of leaves of the marked plant, so as to acquire data of microenvironment parameters that represent growth environments of the leaves.
2. The plant-caring apparatus of claim I, being characterized in that data of the images of the marked plant and the microenvironment parameters related to the leaves collected by the care-related collecting unit (2) and the monitoring module (4), respectively, are simultaneously transmitted to a data-analyzing unit (5), which compares the data of the images of the marked plant against prestored parameter values from a sample database that represent average growth status of the plants in a growth cycle and determines whether the marked plant is wilting, lodging, chlorotic, or yellowing through analysis.
3. The plant-caring apparatus of claim 2, being characterized in that the data-analyzing unit (5) is further loaded with a sample library that contains reference values of the microenvironment parameters that enable the leaves to perform sufficient photosynthesis corresponding to various growth phases in the growth cycle of the plants, so that the data-analyzing unit (5) compares the actual data of the microenvironment parameters related to the leaves of the marked plant collected by the monitoring module (4) against the reference values related to the relevant growth phase and outputs analysis results.
4 A plant-caring apparatus for a plant factory, at least comprising a shuttle cart (1) that moves around a cultivation rack (8) in a cultivation space and a care-related collecting unit (2) that is mounted on the shuttle cart (1) to collect images of plants planted on the cultivation rack (8), the plant-caring apparatus being characterized in that the care-related collecting unit (2) is configured to be carried by the shuttle cart (1) to move around the cultivation rack (8) following a predetermined route while collecting the images of the plants planted on the cultivation rack (8); and the plant-caring apparatus further comprises a processing module (3) for comparing the images of the plants collected by the care-related collecting unit (2) with each other and marking any said plant that exhibits an abnormal condition in a growth status thereof, so that the shuttle cart (1) performs a secondary target-specific patrol and inspection according to marking results, and uses the care-related collecting unit (2) and a monitoring module (4) mounted on the shuttle cart (1) to collect double heterogeneous verification data for a follow-up confirmation operation performed on the marked plant.
5. The plant-caring apparatus of claim 4, wherein the care-related collecting unit (2) collects the data for the follow-up confirmation operation in a way that it is driven by a displacement component (7) to change working positions thereof so as to shift itself among a plurality of observation angles around the marked plant while collecting the images of the marked plant; and the monitoring module (4) changes a working position thereof in a way that it simulates working states of leaves of the marked plant, so as to acquire data of microenvironment parameters that represent growth environments of the leaves.
6. The plant-caring apparatus of claim 5, being characterized in that upon marking of the suspicious plant, the processing module (3) matches the images of the plants with coordinate points along a travel route of the shuttle cart, so that the images of the plants collected in a same time period can be used to determine respective locations of the plants on the cultivation rack (8) through a reverse calculation and a coordinate location of the marked plant on the cultivation rack (8) can be determined.
7. The plant-caring apparatus of claim 6, being characterized in that where analysis results output by the data-analyzing unit (5) according to the data for the follow-up confirmation operation indicate existence of any condition seen in the marked plant that cannot be determined in situ whether it's a disease, the data-analyzing unit (5) transmits data of the images of the marked plant to a control center of the plant factory for further disease interpretation and sample data storage.
8. A plant-caring apparatus for a plant factory, comprising a growth-facilitating device (9), which at least comprises a nutriment loop installed on a cultivation rack (8) to supply nutrients to plants planted on all floating shelves of the cultivation rack (8) through hydroponics or aeroponics, the plant-caring apparatus being characterized in comprising a data-analyzing unit (5) that modulates the growth-facilitating device (9) according to current physical states or characteristics of the plants as collected by a care-related collecting unit (2), so that the growth-facilitating device (9) is enabled to variably supply lighting and nutrient according to growth status of the plants.
9. The plant-caring apparatus of claim 8, being in characterized in comprising a shuttle cart (1) that is formed as a gantry-type structure to be able to span the nutriment loop and has two sides thereof each provided with a lifting unit to uphold a seedling tray placed on one said floating shelf
10. The plant-caring apparatus of claim 9, being characterized in that the shuttle cart (1) is provided with the care-related collecting unit (2) and a robotic gripper for handling the plants, wherein the care-related collecting unit (2) monitors the growth status of the plants, and the robotic gripper performs a planting, thinning or harvesting operation on the plants according to information related to the growth statuses.
11. The plant-caring apparatus of claim 10, being characterized in that after the shuttle cart (1) enters a predetermined range for sensing a said plant, the shuttle cart (1) collects current physical states or characteristics of the plant using the care-related collecting unit (2) through infrared imaging and/or visual imaging.
12. The plant-caring apparatus of claim 11, being characterized in that the nutriment loop comprises photosynthesis nutriment nozzles arranged at a top of each said floating shelf and rhizosphere nutriment nozzles arranged at a bottom of each said floating shelf, wherein the photosynthesis nutriment nozzles release nutriments to stem and leaves of the plants in an aerosol-like form, and the rhizosphere nutriment nozzles release nutriments to roots of the plants in an aerosol-like form.
13. The plant-caring apparatus of claim 12, being characterized in that the shuttle cart (1) is formed as a gantry-type structure to be able to span the nutriment loop and has two sides thereof each provided with a lifting unit to uphold a seedling tray placed on one said floating shelf.
14. The plant-caring apparatus of claim 13, being characterized in that the shuttle cart (1) is further configured to work with a loading/lifting machine to transfer cultivation plates on which the plants are planted.
15. The plant-caring apparatus of claim 14, being characterized in that the loading/lifting machine transfers the cultivation plates to a process unit (10), so that the process unit (10) performs a processing operation on the plants planted on the cultivation plates.