Microsoft developing new robotics platform
Microsoft is developing what it describes as “an end-to-end toolchain to help make it easier for every developer and every organization to create autonomous systems for their own scenarios”.
The tech giant offered a glimpse into one of the components of the new robotics platform at the company’s recent Build conference for developers.
The “limited preview” talks about autonomous systems, motion control, machine teaching, and smart buildings and homes, among other things.
The plan is to enable developers to work with its experts to build intelligent agents using Microsoft’s cloud platform, Azure, and tools such as Microsoft AI.
Developers can also utilize a version of the open-source Robot Open System – specifically, ROS for Windows – through the Microsoft platform.
All these computing tools can be set up to autonomously run physical systems, says Microsoft.
That team building the solution includes longtime Microsoft researchers and engineers and experts from Bonsai, which Microsoft acquired last year.
Bonsai is a Berkeley, California-based startup which has developed a method that abstracts the low-level mechanics of machine learning, so that subject matter experts can more easily train autonomous systems to accomplish tasks.
A key feature of the platform Microsoft highlights is its simulation capabilities, offered in the toolchain in the form of the open-source AirSim.
Originally developed by Microsoft, AirSim was designed to use AI to teach drones, self-driving cars, or robots to learn in simulated environments.
The company says simulation is vital since no real-world factory can afford to make a lot of mistakes, while simulations can be used to analyze thousands or even millions of scenarios in a virtual environment in the cloud.
Ashish Kapoor, Microsoft principal research manager, says: “If I have the ability to spawn thousands of simulations at once and in each one the pedestrian crossing the street is different and the curve of the road is different, suddenly the AI system is able to gather much more diverse experience in a short amount of time.
“Azure gives us the ability to run these simulations at scale, which is really important.”
