Robots that can fit multiple items in a limited space can assist in packing a bag or a rocket to Mars. A team of researchers at the Massachusetts Institute of Technology and Stanford University trained robots using a type of artificial intelligence called “diffusion model” to pack items in a confined space while adhering to a set of constraints: human concerns such as ensuring that heavier items do not crush lighter ones, that some items have a certain amount of space between them, and that the robot’s arm doesn’t collide with the container and damage it, and so on. The researchers say that the diffusion model helped the robots achieve this faster than the training methods used in the past.
Learning Robots Based on Constraints
The learning-based approach aims to allow the artificial intelligence program to learn independently by identifying patterns between training data and desired outcomes. This differs from the rule-based programs previously tested, which are more restrictive as they must behave according to a strict set of programmed rules. Yang says, “The diffusion model is a very good way to select different solutions to a problem while meeting all required constraints.”
Self-Packing Robots Applications
Robots are gaining the ability to pack faster and more efficiently than their human counterparts, with applications that go beyond road trips. The researchers point out that this algorithm could help shipping companies pack diverse items into a single container or assist pharmaceutical companies in delivering a wide range of medications to hospitals in bulk. The possibilities even extend beyond the planet, where a robot could decide how to pack resources if you’re heading to Mars.
Improving Robots’ Decision-Making Capabilities
The M.I.T. and Stanford University team is currently working on making robots more capable of making “sprawled decisions.” This includes not only training the robot to pack within constraints but also training it to do so amid continuously moving variables, for example, when tasked with packing items while moving around a room.
So, the next time you are packing, think of it like a robot to achieve the best results. You might leave the whole thing to machines in the near future.
Leave a Reply