Create a Chip So That Robots Move Better By Working with Humans

Engineers have managed to find the key to make robots more intelligent, at least with a small portion of imagination so they can better calculate their movements.

In an environment where human beings and robots live almost together doing some kind of work such as on an assembly line, both machines and workers cannot be working closely together to avoid accidents.

Advances in artificial intelligence are making it increasingly diffuse that barrier that exists between workers and robots so they can work in better harmony, collaborating with each other and making the accident rate is virtually non-existent.

Now a new processor manufactured by Realtime Robotics allows robots to quickly plan their next moves to keep humans completely safe. This Startup, located in Boston, seems to have hit the key of something that were already looking for many other engineering companies for some time. Realtime Robotics is trying to give robots the kind of low level intelligence needed to move around in a working environment with humans.

Thanks to this chip, when connecting to the 3D sensors of the robots, it allows the machines to consider different effects in all their range of movements, previously imagining the result before choosing the action that best suits the task in question.

In this way, with this processor in a robot arm in an assembly line, the arm before moving to the right side to leave a piece, you can choose to slow down or perform a different movement if the sensors locate the presence of a human in the middle of the road.

The chip is capable of overloading mathematical calculations with a motion planning algorithm developed by Konidaris and other researchers while at Duke University in the United States. By running the calculations in parallel, the processor can perform the movements more than 10,000 times faster than a chip from a normal computer.

This advance has not gone unnoticed by a multitude of robotic arm manufacturers, which is already using Realtime Robotics technology.