Researchers Engineer Low-Cost, Human-Friendly Robot for AI

The robot uses advances in artificial intelligence and deep reinforcement learning to master intricate human tasks.

Researchers at the University of California, Berkeley have engineered a robot ideal for artificial intelligence applications. Named Blue, the machine is low-cost and human-friendly enough to one day be a staple in every home.

RELATED: THIS NEW ROBOT CAN FEED YOU YOUR DINNER

Mastering human tasks

Blue was designed to use recent AI and deep reinforcement advances to learn to master such human tasks as folding laundry or making a cup of coffee.

"AI has done a lot for existing robots, but we wanted to design a robot that is right for AI," said project leader Pieter Abbeel, professor of electrical engineering and computer sciences at UC Berkeley.

"Existing robots are too expensive, not safe around humans and similarly not safe around themselves -- if they learn through trial and error, they will easily break themselves. We wanted to create a new robot that is right for the AI age rather than for the high-precision, sub-millimeter, factory automation age."

Blue's is not only durable but also affordable. In total, the robot costs less than $5,000 to manufacture and assemble.

RELATED: HANSON ROBOTICS' LITTLE SOPHIA TEACHES STEM, ROBOTICS, AI TO CHILDREN

Furthermore, Blue's arms have been ideally designed to be sensitive to outside forces. They can be very stiff or very flexible depending on what a task may require. This is very different from traditional rigid robotics focused on industrial applications. 

Rigidity versus Flexibility

"We've often described these industrial robots as moving statues," said graduate student David Gealy

"They are very rigid, meant to go from point A to point B and back to point A perfectly. But if you command them to go a centimeter past a table or a wall, they are going to smash into the wall and lock up, break themselves or break the wall. Nothing good."

Blue is designed to function in an environment where mistakes are made in order to learn from them. As such, the robot is highly sensitive to feedback, always adapting the amount of force it exerts at any given time.

"One of the things that's really cool about the design of this robot is that we can make it force-sensitive, nice and reactive, or we can choose to have it be very strong and very rigid," added Gealy. 

To achieve these capabilities, the researchers had to decide what features Blue needed and what the robot could do without. As such, they gave Blue joints that can move in the same directions as a human arm but lack some of the strength and precision of a typical robot.

"What we realized was that you don't need a robot that exerts a specific force for all time, or a specific accuracy for all time. With a little intelligence, you can relax those requirements and allow the robot to behave more like a human being to achieve those tasks," said postdoctoral research fellow Stephen McKinley.

Advertisement