While robotics are an increasingly important industry, the AI systems used to robots are hard to measure. But research from UC Berkeley and Google Brain introduced robotics benchmarks for learning with low-cost robots, or the ROBEL benchmark, hoping to solve that problem at a smaller scale.
The ROBEL benchmark was created specifically for testing the AI systems on smaller, lower-cost robotics systems, says a recent VentureBeat article.
Affordability is part of the reason for the research: more affordable platforms means more likely adoption of the technology. And any relatively new field could always use more developers and researchers.
This contrasts the $10,000-$15,000 corporate machines you hear about in blogs like this.
The benchmark aims to measure tasks such as “turn and screw” for D’Claw and “stand and walk” for D’Kitty, says VentureBeat.
“A robotic hand good at screwing or unscrewing things could be used to manipulate valves in a factory or simply to open a jar, while a robot that knows how to get around on four legs can use locomotion to climb over obstacles or travel,” VentureBeat.
Read Next: Walmart Robots: How Should They Work?