US-based robotics firm, Perrone Robotics has announced its plans to integrate a scalable autonomy platform into a full range of robots, vehicles and equipment. The platform, which can run on a computer as small as a Raspberry Pi microcomputer can be integrated into anything requiring a robotic mind.
And we mean anything. From something as small as an automated vacuum cleaner, to a car, even a mining truck that, in some cases, may be as large as a home. Wha’s so unique about the system is that developers only need to input as much power is needed for the particular application, so if it’s for a small robot, it won’t have to pump as much energy through it as, say, a lorry.
The MAX architecture scales uniquely,” company founder Paul Perrone said. “It enables companies at any stage to quickly design and build a broad range of robotic products and applications.”
Although the technology is still only in its early stages, the company is negotiating with a very large car manufacturer and various other tier 1 robotics suppliers to offer the automation platform as their system to control vehicles and other robots, specialising in the driverless vehicle market.
“We do work in industrial mining, automated fork lifts, PC manufacturing robots, [and] robots in the home and in the office,” Perrone said. It’s currently working with Liebherr on its surface mining trucks, preventing them from crashing into obstacles, such as low-lying bridges.
The basis of Perrone’s technology is neural networks, with the idea developing from the way vertebrates learn and move. It’s been created off the back of research by the University of California, San Diego’s Vertebrate Movement Laboratory (VML), which looks at how neuronal calculations develop.
“The [UC San Diego] VML team’s observed data show that almost all of the neuronal calculations occur within sets of neurons within the spinal column,” Perrone expained. “Further, these calculations take on a mathematical form that is entirely different, and completely incompatible with “Deep Learning” approaches that current automotive AI researchers use.”
