MIT creates robot that carefully places food in a container (video)

Researchers from the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT) have presented an intriguing development called RoboGrocery. It's a robotic system that combines computer vision and a soft manipulator to neatly pack a wide range of items, including delicate fruits and food products that are sensitive to compression.

During the demonstration of RoboGrocery's capabilities, researchers placed 10 unknown objects onto a grocery conveyor belt. Alongside solid items like containers of ice cream and boxes of food, there were more compression-sensitive products such as grapes, cakes, and crackers.

The process begins with the computer vision system, which identifies the objects before determining their size and placement on the belt. When the manipulator handles grapes, integrated pressure sensors detect their softness, ensuring they are not placed at the bottom of the bag with other products. The robot then identifies that a can of soup has a more rigid structure, directing it to the bottom of the bag.

"This is an important first step towards robots packing groceries and other items in real-life scenarios. While we're not quite ready for commercial deployment, our research has shown the potential for integrating multiple sensing methods into soft robotic systems," said one of the developers of RoboGrocery, Annan Zhang.

Developers noted there are numerous opportunities to enhance RoboGrocery, including improving grasping techniques and visualization systems to enhance the accuracy of packing order determination. As the system becomes more reliable, it could be used beyond grocery stores, such as in industrial settings and waste recycling plants.