CORES: AI-Based detection of unsecured cargo in the vehicle interior

© Fraunhofer IOSB
Detection of a potentially unsecured object.
© Fraunhofer IOSB
Notification Concept - visible on the vehicle display.
© Fraunhofer IOSB
Concept for an AI-generated warning video.

AI-based in-cabin object detection for greater vehicle safety

CORES (Cargo and Occupant Risk Evaluation System) is a research and demonstration project for detecting and assessing unsecured objects inside the vehicle cabin. The solution extends Fraunhofer IOSB’s AktiMeter technology with functions for in-cabin object detection, unsecured cargo detection, and cargo risk assessment. Camera data is used to detect objects, assess their securing status, and evaluate potential risks. CORES demonstrates how in-cabin cameras can be used not only for driver and occupant monitoring but also for new safety features. The results of the analysis can be transferred to other systems via an interface. A patent application has been filed for the underlying concept.

Unsecured cargo detection and cargo risk assessment in vehicles

Loose objects inside the vehicle cabin can pose a hazard during sudden driving maneuvers or accidents. CORES supports the development of new safety and driver-assistance features by detecting potentially unsecured objects and assessing possible risks.

• Detection of unsecured objects and luggage in the vehicle interior
• Assessment of the securing status and potential risks
• Expansion of occupant monitoring systems and in-cabin monitoring solutions
• Development and investigation of new in-cabin safety features
• Rapid proof-of-concept studies for OEMs and suppliers
• Data output via interfaces for connecting to other systems

For OEMs, Tier 1 suppliers and developers of occupant monitoring systems

CORES is aimed at companies and development teams that wish to explore new in-cabin and safety features based on camera data. The focus is particularly on: vehicle manufacturers and OEM pre-development departments, Tier 1 suppliers for cameras and occupant monitoring, developers of Occupant Monitoring Systems (OMS), in-cabin sensing, UX, and HMI teams, as well as automotive software, computer vision, and AI teams.

New automotive innovations with the AktiMeter prototyping platform

With CORES, new vehicle interior analysis features can be evaluated as customer-specific proof-of-concept demonstrations. The AktiMeter platform provides an existing setup for object detection, safety status assessment, cargo risk assessment, and data output:

• Rapid development and demonstration of a proof of concept
• Modular and customer-specific scope of investigation
• Use of an existing AktiMeter setup for short-notice demonstrations
• Early evaluation of new in-cabin object detection and automotive interior monitoring functions
• Optional technical integration with a customer-provided HMI concept
• Optional user studies following separate coordination

Behavior-based communication with CORES

CORES implements a behavioral science-based communication concept to support load securing in the vehicle. During the loading process, context-sensitive prompts on the vehicle display can indicate appropriate measures for securing the load. Additionally, after the trip, CORES can, upon request, provide an AI-generated video that clearly illustrates how unsecured objects can move around the vehicle interior in the event of an accident.

Note on the development status

CORES is a proof of concept. Communication elements such as safety reminders on the vehicle display or an AI-generated video created on demand are concepts. Their technical implementation and effectiveness require further development and evaluation.

Shape the future: Start using CORES for your vehicle development today

Would you like to evaluate a new in-cabin object detection, cargo risk assessment, or occupant monitoring feature as a proof of concept? Talk to Fraunhofer IOSB about your use case, suitable test scenarios, and a clearly defined project scope.

Application example

 

Department Human-AI ­Interaction

If you'd like to learn more about similar projects, visit the automotive pages of our Perceptive User Interfaces group in the Human-AI Interaction (HAI) department.