Building Shelf-Aware Robots

Whether it is for educational purposes, future job prospects, or just for fun, a lot of people are very interested in learning more about robotics. It is not an especially easy field to jump into, however. The hardware is often very expensive, and expertise in a wide array of topics ranging from mechanical engineering and computer science to electrical engineering is necessary to get very far. All of these difficulties leave many people sitting on the sidelines because they do not know where to even start.

YouTuber LeMaster Tech has just released a project that could be very valuable to individuals finding themselves in this situation. He has walked through the process of designing and building an automated inventory robot that is capable of locating items and performing pick-and-place operations. It is far too simple of a system for real-world use, but it is just right for someone looking to build their knowledge base in the area.

In the world of robotics, the components used in the project may not be especially expensive, but they are not cheap for the casual hobbyist either. The inventory system makes use of an Elephant Robotics myCobot 280 robot arm and an Elephant Robotics myAGV mobile chassis, which will set you back about $800 and $950, respectively. But if you are determined to jump into the field, components such as these will serve you well long after your inventory robot is finished.

LeMaster Tech’s plan was to build a small shelf containing a variety of items, then use the robot to autonomously drive to a user-selected location, pick up the item, then drive it back to its starting position. Toward that goal, the myCobot arm was mounted on top of the myAGV. A proximity sensor was also mounted to the robot, which allows it to determine where items are located on the shelf via 3D-printed parts that hang below each one.

Both the arm and the chassis are powered by their own Raspberry Pi single-board computer. LeMaster Tech programmed each with Python to handle functions like driving, locating items, picking objects up, and putting them back down. GPIO pins on the Raspberry Pis were used for communication between the components, which allowed them to act in sequence — for example, this told the arm to stay still until the robot had driven to the proper destination.

A simple graphical interface to the system was developed for use with a touchscreen display. This allows the user to choose whether the inventory robot should pick up or return an item, and the location on the shelf that is involved in the action.

After starting with a basic robot like this, one could use the same hardware platform to add in more advanced technologies, like artificial intelligence, to make the system more powerful and closer to something that could be used in real-world applications. If you are working on something like that, be sure to let us know all about it!