Researchers say that motion planning software is a huge limitation for robot adoption. Duke University's robotics experts provide a practical solution to this problem by adding a new processor.
Robot custom chip comes out to solve sports planning problems
Now robotics experts from Duke University provide a practical solution to this problem by adding a new processor that can calculate the route that the robot should move. The calculation speed is three orders of magnitude faster than the current method, while the power consumption is only One-twentieth of the current method.
This processor chip is a custom FPGA or field programmable gate array. As the name suggests, these are processors that can be reprogrammed after manufacturing to specifically handle certain tasks. They have been around for decades, but they prove to be very good at problems involving machine learning. For example, Microsoft is using FPGAs for AI cloud services.
The advantages of robots using FGPA are clear. For example, before the robot works, the arm needs to draw the environmental area required for the work. It takes a few seconds to pause and calculate its route. It not only considers from A to B, but calculates the 3D space it occupies there, the so-called "scan volume". With this new chip, the robot arm equipped with FGPA reacts to the new environment almost instantaneously. No need to pause for a few seconds.
Researchers say that motion planning software is a huge limitation for robot adoption. If you can do real-time motion planning, robots can now operate in a dynamic, unstructured environment.
Poker Mat,Poker Table Mat,Round Poker Mat,Poker Pad
Cixi Mingsheng Rubber & Plastic Co.,Ltd. , https://www.popmat.com