The path toward robotic arms capable of performing tasks that require gentle, deft motions is being paved by researchers from Northeastern University in Massachusetts.

To eliminate the jerky and stiff motions of current robotic arms, the researchers relocated the robotic arm's heavy motors from its wrist joints to the base of the machine. This, according to the Northeastern team, made the robotic arm lighter and easier to control and thus capable of better understanding its environment — an obstacle that plagues robot operators controlling robots from a remote location.

Source: Northeastern University

According to researchers, it is difficult to determine where the robot is precisely in terms of its environment — and if it is touching something and how hard it may be touching something, for instance. Consequently, these components influence the robotic arm’s performance and safety.

Using machine learning coupled with real-time information, researchers can now determine how much force is applied so that when the robotic arm grasps or manipulates an object, it does so by mimicking how human muscles sense forces such as how heavy an object is.

The researchers believe that this development could pave the wave for remote surgery or for remotely removing explosive devices to assist bomb dispersal experts.