AI-Based Real Time Location & Recognition System.
Locate as if with your own eyes: the real-time location and recognition system powered by AI object detection and the W2MO Digital Twin — no UWB, no RFID, no tags on the goods.
Real-Time Digital Twin.
- Extension of the W2MO Digital Twin with real-time data
- AI-based recognition units track all objects
- Real-time location, tracking and visualization in the Digital Twin
- Easy integration into existing operations
- Extremely high reliability and stability
- Very low investment and long-term total cost of ownership
Our W2MO Digital Twin provides a real-time location system without UWB, RFID or similar technologies. Recognition units mounted on the ceiling of your warehouse detect and localize every object in the camera view using artificial intelligence.
Each recognition unit consists of a powerful AI computer connected in parallel to 4, 8 or 12 cameras. Each camera has 5 MP resolution and a 120-degree field of view for a wide-angle picture of the warehouse.
The operations area is stored as a true-to-scale Digital Twin in W2MO. From there a camera plan is generated, showing where the cameras and the AI computers should be placed on the ceiling.
When objects move into the view of one of the cameras, they are detected immediately on the GPU of the AI computer — only position data is transferred to the Digital Twin. Knowing the location of every camera in reality and in the Digital Twin lets us transfer each object position from an image to an absolute position in the twin.
Object Localization with AI.
- Pre-trained neural networks recognise standard logistics objects
- Special objects can be added; networks tuned to particular lighting conditions
- Similar-looking objects can be told apart via ID markers
- High accuracy via automatic camera calibration — from 1 mm to 10 cm
Our pre-trained neural networks recognise standard logistics objects such as forklifts, people and pallets in real time. If your operation has special objects, we add them to the networks. Special lighting conditions or other customer-specific requirements can be integrated as well.
The AI detects and localizes every relevant object in an image. For similar-looking objects — e.g. forklifts of the same brand — an ID marker enables unique differentiation. The markers can be mounted quickly on any forklift type.
AI also distinguishes between different states of the same object: whether a forklift is loaded or unloaded, whether a person is wearing a safety vest, and more. Using state-of-the-art techniques we can not only localize an object in the image but also estimate its orientation in space — so the driving direction of a forklift and the lifting height of its fork can be determined, even from a single camera.
Data protection through person pixelation.
- Images are processed directly in the recognition unit
- Detected persons are pixelated automatically
- No personal data is processed
The AI computer in the recognition unit handles all image processing — only position data is transmitted to the Digital Twin.
In support mode the cameras can be configured to expose a video stream. Every detected person is pixelated on the recognition unit itself, avoiding any processing of personal data.
Interested in our camera and AI infrastructure?
Talk to one of our experts — no commitment.