Client is a leading Polyurethane (PU) Foam manufacturer.
The Looper Foam Cutting Machine that client is using at its manufacturing plant can run at a max speed of 120 m/min. However, since the defects (in the form of holes, cuts, patches etc.) are currently being identified by humans, the machine has to be operated at a lower speed. Additionally, there is also a large scope of defects being missed completely which further leads to material being rejected by the customer, thus, increasing the overall cost.
Foam manufacturer needs IoT (Internet of Things)/Machine Learning driven solution to detect manufacturing defects during production stage. The solution will replace the current process of defect detection (involving constant worker oversight) with smart camera device that will capture the video stream every time and pass it onto intelligence system (cloud based) that will identify the defect in foam, if any.
Valiance proposed to build an Intelligent Defect Identification platform where Images/videos (captured through high end camera) of normal and abnormal foams from different stages of production can be submitted to the localized ‘learning service’ that will build analytical models to discern OK vs Not OK characteristics of parts or components of foam that meet quality specifications and those that don’t.
This kind of system can be trained to perform such tasks with a high level of confidence and can be deployed on pre-configured hardware on the factory floor so that there can be very little decision latency during production.
1. Smart Camera Unit
2. Local Processing Unit with Sensors
3. Internet Application to track and monitor defects, view performance statistics
- Train the model by uploading and analysing defect image
- Distribute the trained model to edge systems
- Analyse image & provide result
- View on dashboard