The first self-checkout system was installed at a Kroger grocery store outside Atlanta in 1986. It took several decades, but the technology eventually spread throughout the United States. Given that grocery stores are moving toward automation, robotic packaging likely won't be far behind.
This week, MIT’s CSAIL division will present RoboGrocery, a combination of computer vision and soft robotic grippers that can pick up a variety of items. To test the system, researchers placed 10 objects unknown to the robot on a grocery conveyor belt.
Products ranged from delicate items like grapes, bread, kale, muffins, and crackers to much more sturdy items like soup cans, meal boxes, and ice cream containers. The vision system operates first, detecting objects before determining their size and orientation on the belt.
When the grabber touches the grapes, pressure sensors in the fingers determine that the grapes are actually delicate and should not touch the bottom of the bag. This is something many of us have learned the hard way. Next, I notice that the soup can has a sturdier structure and stick it in the bottom of the bag.
“This is an important first step toward having robots pack groceries and other items in real-world environments,” said Annan Zhang, one of the study’s lead authors. “While not yet ready for commercial deployment, our work demonstrates the power of integrating multiple sensing modalities into soft robotic systems.”
The team notes that there is still plenty of room for improvement, including upgrading the grasper and imaging system to better determine how and in what order items should be packed. As the system becomes more robust, it could be expanded outside grocery stores and into more industrial spaces, such as recycling plants.