SPARCUS TECHNOLOGIES
FREIGHT CONTAINER INSPECTION & SURVEYING
OUR PERSPECTIVE
FREIGHT SHIPPING SURVEYING & INSPECTION
The lack of accurate, timely, and diversified data on freight container/trailer condition and
utilization makes it difficult to make relevant informed proactive decisions.
This invariably leads to a reactive approach, creating inefficiencies and delays, both
operationally and financially.
OUR PERSPECTIVE
FREIGHT SHIPPING SURVEYING & INSPECTION
CLAIM UNFORESEEN
CAPACITY OPTIMIZATION LABOR CONSTRAINTS
PREVENTION/RESOLUTION MAINTENANCE
Liability for container Damage left undetected Inaccurate cargo density A growing shortage of
REGULATORY
damage claims is often REGULATORY
or unfixed leads to costly REGULATORY
estimates result in REGULATORY
skilled inspectors,
passed in closed vicious
COMPLIANCE repairs, write-offs and
COMPLIANCE underutilized trailers,
COMPLIANCE driving up labor costs
COMPLIANCE
loop between shippers, downtime, disrupting wasted fuel, and and making it difficult
carriers and terminals, processing speed and, increased transportation to maintain consistent
increasing delays, disputes consequently, operating costs. inspection quality.
and insurance costs. margins.
EXISTING ALTERNATIVES
Automated Manual
Physical Documentation and
SaaS Sensor-fusion Robotic-assisted
Examination digital reporting
Scalable Holistic Efficient Adaptable to Holistic assessment
assessment unforeseen issues
Efficient Real-time Consistent No special equipment Streamlines data collection
Real-time Complex Suitable for Time consuming Time consuming
integration & complex &
Accessible calibration subtle
Labor intensive May require specialized
damages
Requires training
Minimal specialized High CAPEX Prone to human error and fatigue
CAPEX expertise
Internet Adaptable Not suitable for large-scale, remote or hazardous operations
connectivity
SPARCUS TECHNOLOGIES
AI FOR FREIGHT CONTAINER INSPECTION
Research-driven approach to working with big data.
Development of AI/ML solutions for logistics, transport and supply chain.
HONG KONG 2024 GLOBAL
USE CASES
COMPUTER VISION | DEEP LEARNING
EXTERIOR INTERIOR DENSITY & FILL CARGO ASSET MANAGEMENT
INSPECTION INSPECTION LEVEL CALCULATION RECOGNITION & TRACKING
Mold
Mold
Physical
Physical
API INTEGRATION EXTENDED REALITY 3D RECONSTRUCTION REAL TIME
PLATFORM INTEGRATION
Usage analysis Inspection History
Number of Inspections Weight, capacity & door
Positive/Negative direction
Results Analysis Seal presence
Seal presence Container ID & ISO code
Material & door IMDG label presence &
direction class
Pricing Calculation Inspection Results
Adjusted reporting Raw/Inspection Image
(PDF or damage code Inference Time
e.g. CEDEX/IICL)
*Any Relevant Data Can be Displayed if Requested
MOBILE APP INTEGRATION
(INTERIOR ASSESSMENT)
Demonstrate real-time Receive the real-time
inspection results based on AI streamlined inspected
decisions. container images.
Store the results in the unified Data exchange and
database with the admin interfacing with TOS, ERP and
access. other 3rd-party systems.
VALUE PROPOSITION
Parameter Sparcus Technologies Traditional methods
Speed of inspection (per unit) (per unit)
INSPECTION 3 SEC >3 MIN
REPORT GENERATION 3 SEC >5 MIN
Accuracy (estimated) 95% 70%
Price per inspection (Price per image X number of images)
USA $4.7
HONG KONG $4.3
$1.3
EUROPE (AVERAGE, UNWEIGHTED) $4.2
Hazardous zones Yes Conditionally
Movable objects Yes Conditionally
PRICING MODEL
Total Price = Number of API calls x Monthly Price per API call
Initial Price per API Call Fixed: $1.3
The Price per API Call is updated every month based on last month usage projections.
*API call refers to request made by a user's application to access a specific function or data from a SaaS provider's platform.
VALUE PROPOSITION
Controllable No Vendor Lock-In
Price Points or Dependence
International-level
research expertise
Possibility of Architectural
Customization Scalability
ENGINEERING SUPPORT SERVICE
Support, security, and bug fixing. Access to
general product enhancements.
ROADMAP optionally: Service Level Agreement (SLA) and
continuous AI-Application training
.
8+ MONTH
INTEGRATION & SCALING
Integration of the solution with the user
interface. Scaling the solution across
different modalities, if needed.
6 - 8 month
SERVER
DEPLOYMENT
Creating a publicly available
ALGORITHM server with API endpoints.
DEVELOPMENT
4 - 6 month
DATA ENGINEERING AI model training, validation,
and benchmarking.
Infrastructure & operations review.
Collaborative development,
transition, and lifecycle service offer. 2 - 4 month
Joint use-case review session. 0 - 2 month
Initial data assessment & use-
case mapping. Co-creation
of a consulting offer.
OUR WORK
Development of the leading OCR algorithm for
March handwritten Sino-Korean Characters.
Partners: Deloitte | HKU
Introduction of a new approach for the better
July semantics understanding by a language model.
Partners: HKUST
Development of a new attention mechanism for
Sep Computer Vision tasks, called Smooth Attention.
Partners: HKUST
Development of the leading algorithm for TSP
Oct fundamental logistics problem based on AI.
Partners: IE University
OUR TEAM
Boris Kriuk Fedor Kriuk Artem Alkhamov Wolfgang Lehmacher Phillipe Brochard
Chief Executive Officer Chief Technical Officer Chief Commercial Officer Advisor Advisor
5 Int. Olympiad Medals 5 Int. Olympiad Medals 6 years in B2C sales Ex-CEO of GeoPost, Ex-CEO of Auchan France,
Ex-Director of World Ex-Chief Restructuring Officer
3 Scientific Publications 3 Scientific Publications 5 years in B2B sales
Economic Forum (WEF)
2+ years in ML Consulting 3+ years in Software Dev Retail, Procurement, Strategic
2+ years in warehouse logistics
Supply Chain, Logistics, Planning
ML, DL, Data Science, ML, Data Science, Full- Marketing, International Transportation, Strategic
MLOps, Physics Stack SWE, Physics Economics Planning
www.sparcus-technologies.com +852 9162 6635 outreach@sparcus-technologies.com Suit C / Level 7, 50 Stanley Street, Hong Kong
