WO2022200234A1

WO2022200234A1 - Use of devices with inbuilt accelerometers to detect vibration on board a rail or road vehicle

Info

Publication number: WO2022200234A1
Application number: PCT/EP2022/057240
Authority: WO
Inventors: Alexander Roger David BRAIN
Original assignee: Alonyx Ltd; Pandrol Vortok Ltd
Current assignee: Alonyx Ltd; Pandrol Vortok Ltd
Priority date: 2021-03-22
Filing date: 2022-03-18
Publication date: 2022-09-29
Anticipated expiration: 2023-09-22
Also published as: GB202103936D0; GB2606335A

Abstract

Embodiments of the invention relate to the use of devices with inbuilt accelerometers, for example smartphones, to detect vibration on board a rail or road vehicle. For example, according to one embodiment there is provided a method of remotely identifying a damaged area on a surface of a rail head or a road based on acceleration data collected during at least one journey of a vehicle over a section of rail track or road, wherein: (a) a damaged area is identified based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or (b) a damaged area is identified based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or (c) a damaged area is identified based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data.

Description

USE OF DEVICES WITH INBUILT ACCELEROMETERS TO DETECT VIBRATION ON BOARD A RAIL OR ROAD VEHICLE

Technical Field

The present invention relates generally to the use of devices with inbuilt accelerometers to detect vibration on board a rail or road vehicle.

Background

Some consumer devices, such as smartphones, smartwatches and tablets, have inbuilt accelerometers. The data from such accelerometers may be collected by an application on the device (for example, a fitness application) and processed, either on the same device or on another device in communication with it, to provide information to the user of the device.

It has been proposed that accelerometer data from smartphones carried by train passengers could be used to obtain a measure to assess vibration-based train ride comfort or track performance. It has also been recognised that accelerometer data from a smartphone on board an underground train may be compared to stored data to identify where on an underground train line a passenger carrying the smartphone is located.

It is desirable to provide improved methods of utilizing data generated by an accelerometer on a consumer device, or another suitable device, to detect vibration or movement within a vehicle carrying the device that is caused by a railway track or road surface over which the vehicle is travelling.

Summary of Invention

According to an embodiment of a first aspect of the invention there is provided a method of assessing vehicle ride quality in respect of a section of rail track or road based on acceleration data collected during at least one journey of the vehicle over the section of rail track or road, wherein: (a) assessment of vehicle ride quality is based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or (b) assessment of vehicle ride quality is based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or (c) assessment of vehicle ride quality is based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data. For example, the vehicle may be a train, tram, bus or coach.

According to an embodiment of a second aspect of the invention there is provided a method of remotely monitoring condition of a rail based on acceleration data collected during at least one journey of a vehicle over a section of rail, wherein: (a) condition of the rail is assessed based on acceleration data relating to a journey of the vehicle over the section of rail obtained from more than one source; and/or (b) condition of the rail is assessed based on acceleration data relating to more than one journey of the vehicle over the section of rail; and/or (c) condition of the rail is assessed based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data. For example, the vehicle may be a train or tram.

In a method embodying the second aspect of the invention, condition of the rail may be assessed using acceleration data collected during a plurality of journeys made over a period of time, in order to monitor changes in rail condition over that time period.

According to an embodiment of a third aspect of the invention there is provided a method of remotely identifying a damaged area on a rail, for example a surface of a rail head, or a road based on acceleration data collected during at least one journey of a vehicle over a section of rail track or road, wherein: (a) a damaged area is identified based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or (b) a damaged area is identified based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or (c) a damaged area is identified based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data. For example, the vehicle may be a train or tram.

According to an embodiment of a fourth aspect of the invention there is provided a method of assessing passenger ride comfort during a passenger vehicle journey in respect of a section of rail track or road based on acceleration data collected during at least one journey of the vehicle over the section of rail track or road, wherein: (a) assessment of passenger ride comfort is based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or (b) assessment of passenger ride comfort is based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or (c) assessment of passenger ride comfort is based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data. For example, the vehicle may be a train, tram, bus or coach.

In a method embodying an aspect of the invention, the acceleration data may be provided by sensing means located in a portable computing device carried on the vehicle. The portable computing device may be a consumer device, such as a smartphone, a tablet or a smartwatch. The sensing means may comprise an accelerometer.

When recording acceleration data, the sensing means may be located at any position on the vehicle, for example at a location in the driver’s cab or at a location associated with any passenger seat, or at several positions on the vehicle, for example at multiple locations associated with passenger seats of the vehicle.

A method embodying an aspect of the invention may further comprise collecting measurements of angular rotational velocity during the or each journey, and using the angular rotational velocity measurements in conjunction with global positioning system data to estimate the location of the vehicle. The method may further comprise collecting measurements of magnetic field during the or each journey, and using the magnetic field measurements in conjunction with the angular rotational velocity measurements and global positioning system data to estimate the location of the vehicle.

In a method embodying an aspect of the invention, the acceleration data may be analysed using a machine learning algorithm.

According to an embodiment of a fifth aspect of the present invention there is provided a computer program which, when run on a computing device, causes that computing device to carry out a method embodying the first, second, third or fourth aspects of the invention.

According to an embodiment of a sixth aspect of the present invention there is provided a suite of computer programs which, when run on one or more computing devices, causes those devices to carry out a method embodying the first, second, third or fourth aspects of the invention.

Brief Description of Drawings

Figure 1 is a block diagram of a smartphone for use in carrying out a method embodying the present invention;

Figure 2 shows a view of the smartphone of Figure 1 when used in an embodiment; Figure 3 is a diagram illustrating processing of data by an algorithm on a device; Figure 4 is a diagram illustrating inputs and outputs of a neural network used by the algorithm; and

Figure 5 is a diagram illustrating processing carried out on a device and online (“off- device”).

Detailed Description

Some embodiments relate to assessing vehicle ride quality, for example the ride quality of vehicles that run on tracks such as trains or trams, or of road vehicles such as cars, vans, buses, coaches or lorries.

Some embodiments relate to remote monitoring of railway rail condition, for example so as to monitor deterioration in the condition of a railway rail over time.

Some embodiments relate to remotely identifying areas of damaged rail track or road surface, for example areas of excessive rail head wear or pot-holes in a road surface. For example, sections of rail along a railway track which are of poor quality and may be in need of attention may be located, without the need first for a dedicated inspection vehicle or trained person to go along the track to do this (although a person/inspection vehicle may be required subsequently to verify the condition of a rail at the section located). Some embodiments relate to assessing train ride quality. In some cases, assessment of train ride quality may be a way in which railway rail condition may be monitored remotely. Some embodiments relate to assessing passenger ride comfort during a passenger vehicle journey.

In an embodiment it is proposed to use data from accelerometers built into one or more consumer devices, such as smartphones, smartwatches (or other body-worn devices) and tablets, to detect vibration or movement caused by a railway track or road surface over which a vehicle carrying the device is travelling. For example, a device configured with software for collecting the required data may be positioned on a table, desk or other surface on board a vehicle. In the case of a body-worn device which is normally retained on the body whilst measurements are collected, such as a smartwatch, it may be necessary to apply appropriate processing to discern between movements of the body (e.g. wrist or arm) and the vehicle itself. In some circumstances, it may be advantageous to collect acceleration measurements from both a device worn on a person’s body and a device positioned on a table or the like (preferably close to the person), as the difference between the measurements may give an indication of the response of the human body to vibrations experienced during a vehicle journey.

As a vehicle travels along, the acceleration values along one, two or three axes may be recorded. Sensor software may be configured to report data only when there is a change in what the sensor has detected, or may be configured so that sensor measurements are sampled regularly. For example, the acceleration data may be sampled at regular time intervals, such as 50Hz.

From this accelerometer data, a computer algorithm (on the device or elsewhere) may determine, for example, one or more peak acceleration values, which may denote track/road areas of concern. Alternatively, or in addition, a wider measure of rail track/road quality and/or passenger comfort may be derived from the data. For example, filtering may be applied in line with existing standards for passenger vehicle vibrations (such as ISO 2631-5:2018, which relates to methods for evaluating human exposure to multiple mechanical shocks, or BS EN 12299:2009, which relates to methods for quantifying the effects of vehicle body motions on ride comfort for passengers and vehicle assessment with respect to ride comfort), with a view to deriving some kind of ride quality index. The software on the device may be an app running on the device in the background and hence be effectively invisible to users (unless they wish to add comments -see later).

Subject to security and privacy issues being dealt with appropriately, it may be possible for data to be gathered from an app (application/computer program) running on devices of passengers on a vehicle. This may provide multiple sets of data, possibly tens or hundreds, from locations which may be spread out throughout the vehicle. However, it may be a challenge to persuade passengers in general (i.e. members of the public) to download the app onto their devices and use the app to record the data. A suitable alternative might be to require or induce one or more staff employed by the organisation responsible for maintenance of the rail track or road under inspection, or for running the passenger transport (e.g. trains, trams or buses) on that rail track or road, such as a train or bus driver, guard, conductor or another who works on board the vehicle whilst it is travelling, to download the app and run it during their working day as the vehicle travels to record the data needed.

One embodiment of the software records data from sensors on devices such as smartphones, smartwatches or tablets used by one or more train staff travelling on board a train during a journey to obtain data from which to assess train track quality.

For example, the data may be used to assess whether the track quality is sufficiently bad as to require remedial work and/or whether remedial maintenance that has been carried out on a section of track has had the intended beneficial effect. Using the devices of train staff may allow the software to be deployed more easily, with fewer approvals needed, as the commercial benefit to bodies responsible for track maintenance will be clearly understood.

It may be desirable to aggregate data from two or more, preferably multiple, devices in order to provide a more robust assessment of ride or track/road quality, and/or any change in that quality over time. For example, data from more than one device on board a train during the same or a different journey along the same section of track may be combined to provide a broader range of data from which conclusions may be drawn.

In some applications it will be desirable to correlate accelerometer data obtained from one device with accelerometer data from one or more other devices, or to correlate accelerometer data from one or more devices with rail track or road position, using position data recorded alongside the accelerometer data. Whilst consumer devices are often provided with a satellite global positioning system (GPS), this may not always work reliably when the device is within a vehicle. For example, train windows often provide a very limited, and sometimes filtered, view of the sky so the strength of the satellite signal is reduced. In addition, as satellites may not be evenly distributed across the sky, the accuracy of position information from the GPS may have an unsuitable lower limit.

To assist with better locating where accelerometer measurements have been taken, angular rotational velocity data from a gyroscope and/or magnetic field data from a magnetometer provided on the device(s) may also be recorded, in addition to the accelerometer data and associated GPS data, and used to estimate the position of the vehicle on the rail track or road. For example, gyroscope data may be utilised to obtain a more precise location of the vehicle along the rail track/road.

Gyroscope and/or magnetometer data may also be used with the measured accelerometer data itself to correlate the accelerometer data with other accelerometer data recorded on another day or days, or from one or more other devices.

Figure 1 is a block diagram of a smartphone 1 suitable for running a computer program (app) embodying an aspect of the present invention. In addition to having a display screen 2, inside the smartphone there is a computer processor 3 connected to a memory 4, for storing data used and/or generated by the processor 3, and a communication module 5 including an antenna (not shown) for communicating with, for example, mobile phone networks and a satellite global positioning system (GPS). Communication module 5 includes a GPS unit 50 for determining GPS coordinates.

The smartphone also includes three sensors: an accelerometer 6 which detects and measures acceleration in three mutually perpendicular directions, a gyroscope 7 which detects and measures angular rotational velocity, and a magnetometer 8 which detects and measures magnetic fields to determine direction. Processor 3 is configured to run one or more computer programs for collating and/or processing data from sensors 6, 7 and/or 8, and/or displaying collated and/or processed data or other information on display screen 2, and/or sending processed or unprocessed sensor and/or other data to other computing devices via communication module 5. Data used by the processor 3 may be stored outside the smartphone 1 , for example on internet/cloud storage, in addition to or instead of in memory 4.

It may be advantageous if the data recorded includes comments on ride quality or ride experience provided by a user of the device at one or more points during the journey, especially if the user is a person skilled in assessing rail track/road quality, performing rail track/road maintenance or the like. For example, a person may type in one or more comments, or record a continual or sporadic audio commentary, as the journey progresses, either on the general ride quality/experience or periods of special note (for example, a period when the ride is particularly bumpy, smooth or uncomfortable).

Figure 2 shows smartphone 1 with display screen 2 showing a touchscreen keyboard 21 and comment box 22 suitable for this purpose. If the device has a voice recognition (“hands-free”) capability, this may advantageously be used by the device user (e.g. the vehicle driver) to control the app and/or make comments more easily and without significant distraction.

Some or all of the analysis of the data may be carried out by a machine learning algorithm, such as one or more neural networks trained to recognize and classify acceleration and/ or other input data. The aforementioned recorded user input (written and/or spoken) may also be used as input to the machine learning algorithm. During or before analysis of the recorded data it may be processed to exclude any data which is not needed and/or useful.

In one example, an assessment of track or ride quality or comfort derived from analysis of the recorded data may be presented as an index number indicative of track or ride quality/comfort that is determined using the recorded data at time or distance intervals. Alternatively, or in addition, the assessment or processed data may be presented as a time history chart, a bar chart or in any other suitable format. In another example, processed data for rail condition monitoring or rail/road damage location may be presented as a chart showing peak or characteristic vibrations over the length of the rail/road inspected.

In addition to recording data for analysis, the software on the device may include one or more algorithms for analysing all (or only samples) of the recorded data, so that assessment of ride/track quality may also be carried out on the device itself. A description of an algorithm, which may be used by the app when loaded on a smart device (smartphone, tablet, etc.) for recording data on board trains, will now be provided. In this example the app may record data from the following sensors:

• A 3-axis accelerometer

• A 3-axis gyroscope

• A 3-axis magnetometer

• A Global Navigation System (GNS)

Users may also add notes or annotations to the data recording on the device, enriching the recorded data with their expertise and local knowledge in real-time.

Recorded data will be available on the smart device for subsequent review and may also be automatically uploaded to a cloud-based storage system for off-line viewing and analysis.

The app may indicate (e.g. display) to a user only the recorded data from the device’s sensors, and any user annotations, but optionally may provide further relevant information.

Figure 3 shows blocks of processing of the data which are performed on the smart device, e.g. smartphone 1 , itself (hereafter also referred to as “on-device”):

• Sensor Data 10 - The data 10 from the sensors (e.g. sensors 6, 7 and/or 8 and/or GNS) is stored in the device’s memory 4 (or other in-built storage), for later review and optionally also for transmission to cloud storage. Any user- provided input 100 may also be stored alongside the sensor data 10. The sensor data 10 may be displayed graphically to the user.

• Spatial Alignment 11 - As the device measuring the data can be positioned in any orientation, the sensor data 10 is “spatially aligned” by rotating the 3-axis measurements from the accelerometer, gyroscope and magnetometer so that the measured axes are aligned with the train and track. To do this, Principle Component Analysis (PCA), for example, may be used to find the vertical axis from the acceleration due to gravity, and the longitudinal axis aligned with the direction of travel of the train. The lateral axis, across the train, is then simply orthogonal to the other two axes.

• Temporal Alignment 12 - Off-line temporal alignment of the recorded sensor data 10 is relatively straightforward. For example, two journeys between known locations may be compared, with the GNS locations and speed profile used to dynamically adjust the timings associated with the data points so that the two journeys not only take the same time but are aligned throughout the journey. A Discrete Dynamic Time Warping (DDTW) algorithm may be used for this. Temporal alignment of the recorded data on the device may take the form of applying a correction between the speed at which the measurements were made, and the line speed, or highest recorded speed. A degree of correction may be achieved, for example, using a simple scaling factor, but a more accurate correction may be obtained by using a frequency spectrum adjustment, based on comparisons of spectral content across multiple journeys at different speeds. A modular neural network may for example be used to determine the frequency spectrum adjustment. Journey specific training of one module of such a neural network will determine the response of the vehicle at the particular measurement position for the particular orientation. Another module of the network will be trained to map this vehicle’s response to the more generic response to previous journeys along the same route, which will allow for compensation for this vehicle and measurement speed. Training of this second module of the network may be based only on previous run data stored on the particular device, but may include data from other devices for this route. That is, a frequency spectrum adjustment necessary to achieve temporal alignment of the recorded data may be determined based on the output of a modular neural network trained using data recorded over the same route by both the particular device and other devices.

• Neural Network 13 - A trained neural network 13 for analysing the data may also use a modular approach, with a number of different neural network models used to arrive at the results. Initial training of the neural network models may be done by an external system, then the models may continue to learn from there by downloading new training data 40, derived for example from “off- device”, i.e. external, analysis 160 of datasets collected by one or more other devices. The diagram shown in Figure 4 provides an overview of the inputs and outputs of the neural network 13: o Inputs - In this example, the sensor data 10, along with the spatially aligned (11) and temporally aligned (12) versions of the data described above, are all used as inputs to the models. Desirably this is accompanied by input 100 from a user on board the train, since this user input 100 may provide “semi-supervised” real-time updating of the classification networks. The user input 100 may be obtained by the user simply clicking on a button on the device to, for example, directly identify the section of track as poor quality, or causing rough ride, or going over a crossing or a bridge perhaps. The network models may use this “direct” classification as a training input, for updating the model, alongside the sensor inputs. Optionally, users may also enter “free-text” data, e.g. notes about which train type or vehicle they are on, or where the train is going, etc (for example, either by typing in the information or using speech to text capabilities of the smart device). A pattern recognition network may be applied to this “free-text” input to extract track classifications. Training of such a pattern recognition network may be carried out “off-line” using data (which may be labelled) from multiple devices, if necessary with a language dependency. The neural network models 13 may also be updated using training data derived externally by an “off-device” processing system 160. Outputs - In this example two sets of outputs from the network models are displayed to the user, namely ride quality index or indices 131 and track classification 132.

^■ The notion of a ride quality index is well defined in the vehicle industry, however the neural network models used in the proposed algorithm allow for a ride quality index 131 to be calculated from data recorded by one or more smart devices located on or near a passenger(s) or driver who is generally in a seated position, and to account for the effects of train speed and position. The ride index model may be trained off-line relatively easily on synthetic datasets with a known ride index calculated using a more traditional method, then the model(s) may be applied to the real-time data.

^■ Track classification 132 has two aspects:

The “arrangement” of the track, i.e. whether “plain line”, in a curve, or a switch or a crossing, and The quality of the track, which is primarily whether the track is of poor quality, or could produce a rough ride.

Two different classifier networks may be used, with the output of a track arrangement classifier being used as an input to a ride quality classifier. As well as using gyroscope and GPS data to locate curves and straight sections, training of the track arrangement qualifier may include “off-line” data from other devices classified into straight, curved or switch/crossing datasets. Training of the ride quality network may be in a “semi-supervised” manner, in that user inputs may be used to identify if the ride is rough or of poor quality, as well as using data from other devices to initially train or update the network.

As mentioned above, and as illustrated in Figure 5, the algorithm may have access to additional training data 40 provided by an “off-device”, i.e. external, system 160, which analyses data 150 from multiple external sources to extract additional training data 40 and then provides the additional training data 40 to a training data store on the device(s). The additional training data 40 may be used as input to update the neural network 13, to assist with identifying and classifying ride features, and also potentially to notify users of approaching track features or events. Such additional training data 40 may, for example, include comments that other users have made on the same section of track or other relevant ride data already collected, but may also include features of interest extracted by the software for comparison purposes.

As described above, methods embodying the present invention may be carried out on a computing device such as a smartphone, for example like that illustrated in Figures 1 and 2, or another computing device such as a tablet, smartwatch or the like. Such computing devices need not have every component illustrated in Figures 1 and 2, and may be composed of a subset of those components.

The above-described embodiments of the present invention may advantageously be used independently of any other of the embodiments or in any feasible combination with one or more others of the embodiments.

Claims

1. A method of assessing vehicle ride quality in respect of a section of rail track or road based on acceleration data collected during at least one journey of the vehicle over the section of rail track or road, wherein:

(a) assessment of vehicle ride quality is based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or

(b) assessment of vehicle ride quality is based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or

(c) assessment of vehicle ride quality is based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data.

2. A method of remotely monitoring condition of a rail based on acceleration data collected during at least one journey of a vehicle over a section of rail, wherein:

(a) condition of the rail is assessed based on acceleration data relating to a journey of the vehicle over the section of rail obtained from more than one source; and/or

(b) condition of the rail is assessed based on acceleration data relating to more than one journey of the vehicle over the section of rail; and/or

(c) condition of the rail is assessed based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data.

3. A method of remotely identifying a damaged area on a rail or a road based on acceleration data collected during at least one journey of a vehicle over a section of rail track or road, wherein:

(a) a damaged area is identified based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or

(b) a damaged area is identified based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or

(c) a damaged area is identified based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data.

4. A method of assessing passenger ride comfort during a passenger vehicle journey in respect of a section of rail track or road based on acceleration data collected during at least one journey of the vehicle over the section of rail track or road, wherein: (a) assessment of passenger ride comfort is based on acceleration data relating to a journey of the vehicle over the section of rail track or road obtained from more than one source; and/or

(b) assessment of passenger ride comfort is based on acceleration data relating to more than one journey of the vehicle over the section of rail track or road; and/or (c) assessment of passenger ride comfort is based on comments provided by at least one person travelling on the vehicle while the acceleration data was collected, in conjunction with the acceleration data.

5. A method as claimed in any preceding claim, wherein the vehicle is a train or tram.

6. A method as claimed in claim 1 or 4, wherein the vehicle is a bus or coach.

7. A method as claimed in any one of claims 1 to 6, wherein the acceleration data is provided by sensing means located in a portable computing device carried on the vehicle.

8. A method as claimed in claim 7, where the portable computing device is a consumer device.

9. A method as claimed in claim 8, wherein the consumer device is one of a smartphone, a tablet and a smartwatch.

10. A method as claimed in any one of claims 7 to 9, wherein the sensing means comprises an accelerometer.

11. A method as claimed in any preceding claim, further comprising collecting measurements of angular rotational velocity during the or each journey, and using the angular rotational velocity measurements in conjunction with global positioning system data to estimate the location of the vehicle.

12. A method as claimed in claim 11 , further comprising collecting measurements of magnetic field during the or each journey, and using the magnetic field measurements in conjunction with the angular rotational velocity measurements and global positioning system data to estimate the location of the vehicle.

13. A method as claimed in any preceding claim, wherein the acceleration data is analysed using a machine learning algorithm.

14. A method as claimed in claim 13, when read as appended to claim 7, wherein the machine learning algorithm uses additional data from at least one source external to the portable computing device when analysing the acceleration data.

15. A method as claimed in claim 2, wherein, to monitor changes in rail condition over a period of time, condition of the rail is assessed using acceleration data collected during a plurality of journeys made over that time period.

16. A method as claimed in claim 7, wherein, before the acceleration data is used for assessment, monitoring or identification purposes, it is spatially aligned by rotating measured axes of the data with the vehicle and the rail track or road.

17. A method as claimed in any preceding claim, wherein, when a set of acceleration data from one journey along a route is to be compared with a set of acceleration data from another journey along the route, the two sets of acceleration data are temporally aligned by applying a frequency spectrum adjustment.

18. A computer program which, when run on a computing device, causes that computing device to carry out the method of any preceding claim.

19. A suite of computer programs which, when run on one or more computing devices, causes those devices to carry out the method of any of claims 1 to 17.