Smart TErrestrial WArehouse Robot Deployment
Companies manufacturing, processing, and selling steel pipes have to store these huge pipes in large outside yards. As a result, they face many difficulties in inventory management. These difficulties majorly include:
- Counting the number of pipes in a particular stack in the inventory
- Inability to automate the counting process and dependency on manual labor leads increase in cost
- Keeping a live record of pipe data like count, material, diameter location in inventory, etc.
A generic ROS based rover stack which is capable of going to periodic pipe counting missions. Rover will be equipped with sensors like a camera, GPS, etc. that will be used for all the labor-intensive tasks like going in the field and counting. Rover has a computer vision algorithm deployed that will be used to calculate the number of pipes in a stack using the camera’s video stream. Rover is also responsible to publish useful information on Grafana dashboards like:
- Count
- Pipe manufacturing type, outside diameter and material
- Rover health stats like current drawn, charge estimate, power monitoring, etc.
- ROS
- Husky Rover by Clearpath Robotics
- OpenCV
- Python
- Influx DB
- Grafana
- The solution is highly customised as it uses ROS stack due to which it easy to port to any autonomous solution. This makes our rover stack a plugin like entity that can be deployed to any autonomous solution including Rovers, drones etc.
- An intuitive dashboard which increases observability about:
- Count of pipes for each rover mission
- Pipe material, manufacturing type & diameter
- Rover Stats like current drawn, battery
- Separation of concerns between rover stack, computer vision and dashboards which makes our solution modular and customizable.
- Makes use of Terrestrial Rover (Husky) instead of drones which provides following benefits:
- Drones are more susceptible to weather changes like rain whereas rover provide more robust and reliable solution.
- Due to their high payload capacity, rovers will be more flexible and customizable. They may also come in handy for other use cases.
Bhavya hoda, Prabhsimar singh Taneja, Jasmeet Singh, Amanjeet Singh