Vision AI Pilot - Germany - Manufacturing
Budget: €750 – €1,500 EUR
Pilot Project
In a Manufacturing plant, a pilot will take place using Vision AI for Machine Condition Monitoring.
The setup:
1. Robot - MIR with Cobot Arm
2. Hi-Res camera
3. AI solution for Image processing and model build/training
Plan:
Robot will do inspection rounds around 60 sock production machines. Each time, the robot will dock at a machine and makes photos or videos of 5 inspection points. It will around 2-3 minutes and the robot will move on to the next machine. The images are processed through the AI model. Feedback on anomalies is sent via notifications and dashboards.
Objective
Over time the machine anomalies can be identified based on the images captured and preventive steps can be taken to avoid machine downtime/breakdowns.
Scope:
- Capture images from Camera
- Annotate and Classify images
- Select and Deploy AI model
- Train model
- Run the Images through
- Provide the feedback from anomalies
Activities:
- Setup image transfer from camera to cloud
- Setup the cloud AI solution with Image processor
- Setup cloud AI model(s) (YOLO)
- Train model with images
- Execute and collect Condition monitoring feedback
Out of scope:
- image classification and annotation will be done by other team
- Robot and HiRes camera is already covered and not needed in this scope.
Other:
For pilot, first 1-2 weeks on site to make solution run.
After pilot: build the Industrial Ready solution, with EDGE AI device, tuned model, maintain/optimize
Location: Western Germany, close to Duisburg
Ideal skills and experience:
- python/pytorch
- AI tech and modeling
- Dashboard feedback: gecko board/klipfolio
- Experience in industrial environments (manufacturing shopfloors)
Looking for proposals with minimal setup for Pilot (cost efficient) to prove the realistic machine condition feedback. Then we need to build the scalable production ready solution with all bells and whistles. This solution can be used for other factories and other clients.
Pls only European Residents
In a Manufacturing plant, a pilot will take place using Vision AI for Machine Condition Monitoring.
The setup:
1. Robot - MIR with Cobot Arm
2. Hi-Res camera
3. AI solution for Image processing and model build/training
Plan:
Robot will do inspection rounds around 60 sock production machines. Each time, the robot will dock at a machine and makes photos or videos of 5 inspection points. It will around 2-3 minutes and the robot will move on to the next machine. The images are processed through the AI model. Feedback on anomalies is sent via notifications and dashboards.
Objective
Over time the machine anomalies can be identified based on the images captured and preventive steps can be taken to avoid machine downtime/breakdowns.
Scope:
- Capture images from Camera
- Annotate and Classify images
- Select and Deploy AI model
- Train model
- Run the Images through
- Provide the feedback from anomalies
Activities:
- Setup image transfer from camera to cloud
- Setup the cloud AI solution with Image processor
- Setup cloud AI model(s) (YOLO)
- Train model with images
- Execute and collect Condition monitoring feedback
Out of scope:
- image classification and annotation will be done by other team
- Robot and HiRes camera is already covered and not needed in this scope.
Other:
For pilot, first 1-2 weeks on site to make solution run.
After pilot: build the Industrial Ready solution, with EDGE AI device, tuned model, maintain/optimize
Location: Western Germany, close to Duisburg
Ideal skills and experience:
- python/pytorch
- AI tech and modeling
- Dashboard feedback: gecko board/klipfolio
- Experience in industrial environments (manufacturing shopfloors)
Looking for proposals with minimal setup for Pilot (cost efficient) to prove the realistic machine condition feedback. Then we need to build the scalable production ready solution with all bells and whistles. This solution can be used for other factories and other clients.
Pls only European Residents