US robot packs groceries with record accuracy

An automated soft robot has achieved remarkable precision in packing groceries. Developed by researchers, it significantly reduces errors compared to traditional methods, demonstrating its potential to revolutionize grocery packing operations with enhanced efficiency and accuracy.

Researchers from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) have created a new robot specifically designed to automate the task of bagging items in grocery stores.

This innovative soft robotic system incorporates soft-touch sensors, motor-driven proprioception, advanced vision technologies, and a novel algorithm, Caliber.Az reports citing the foreign media.

Named RoboGrocery, this system is capable of efficiently handling a continuous stream of diverse items on a conveyor belt. The soft robotic arm equipped with multiple sensors accurately assesses the size and firmness of each object. This capability allows it to translate the abstract concept of a "well-packed container" into precise and achievable objectives, marking a significant advancement in robotic packing technology.

Conventional robotic bin-packing has typically centered on handling rigid, rectangular items, often encountering difficulties with objects of varying shapes, sizes, and stiffness.

"The challenge lies in making rapid decisions about whether to pack each item, particularly since we do not make any assumptions about the object as it arrives on the conveyor belt," remarked Annan Zhang, the lead author of the study and a Ph.D. student at MIT CSAIL, in a statement published on The Robot Report.

MIT's RoboGrocery surpasses these challenges through an innovative integration of RGB-D cameras, closed-loop control servo motors, and soft tactile sensors. The RGB-D cameras capture both depth and color data, facilitating precise identification of object shapes and sizes as they move along the conveyor belt.

The researchers explain that the servo motors ensure precise control and feedback, enabling the gripper to dynamically adjust its grip according to the specific properties of each object. This sophisticated combination enhances the robot's ability to handle a wide range of items effectively and efficiently.

Moreover, the gripper's integrated sensors are capable of measuring pressure and deformation, providing valuable information about an object's stiffness and fragility. This advanced capability enables RoboGrocery to handle a wide variety of items effectively.

For precision item sorting, researchers conducted tests to evaluate RoboGrocery's effectiveness. They randomly placed ten different realistic grocery items on a conveyor belt and repeated this process three times. The evaluation focused on determining how frequently heavy items were placed on top of delicate ones.

MIT CSAIL's report highlighted that the soft robotic system demonstrated significant superiority over traditional methods. It performed nine times fewer actions that could potentially damage items compared to a sensorless baseline utilizing pre-programmed motions, and 4.5 times fewer than a vision-only approach lacking tactile sensing.

In a practical example, when grapes and a can of soup were conveyed, the RGB-D camera accurately identified and measured them. The gripper gently handled the grapes, while tactile sensors recognized their delicacy and placed them in a designated area. Subsequently, the gripper efficiently managed the soup, identified as sturdy, and directly placed it in the bin. Finally, the grapes were delicately positioned on top, showcasing the robot's precise and careful handling capabilities.

According to researchers, the system, controlled by a microprocessor, exhibited strong performance across a variety of items such as bread, chips, soup cans, and ice cream containers.

Despite its success, there is still room for improvement. The researchers acknowledged that the current method for assessing an item's delicacy is relatively rudimentary.

"Our current grasping techniques are quite basic, but enhancing these methods could yield significant advancements," explained Zhang. "For instance, determining the best grasp direction to reduce failed attempts and effectively manage items arriving on the conveyor belt in challenging orientations."

Zhang elaborated, "A cereal box placed flat may be too large to grasp from above, but if it's standing upright, it could be much easier to handle."