Realtime Robotics Inc., a Boston-based startup with chip technology that enables industrial robots to operate more safely and efficiently, has raised $31.4 million in fresh funding to support growth initiatives.

The startup announced the Series A round today. The capital, Realtime Robotics says, was provided by a group of more than a half-dozen backers that included Toyota Motor Corp., industrial automation technology providers and investment firms.

Robot arms in factories perform tasks such as assembling electronic appliances through a process known as motion planning. The computer powering each robot analyzes the parts from which it will assemble the appliance and comes up with several different ways of carrying out the task. It then identifies the most efficient approach by weighing various operational factors, such as the number of components that need to be cobbled together. 

Realtime Robotics offers a hardware module based on a custom chip that it claims can perform motion planning 1,000 times faster than traditional algorithms running on graphics cards. The startup’s website contains few details on how its technology achieves the speedup. However, Realtime Robotics did divulge a few technical details about its approach in 2019 to IEEE Spectrum, a magazine operated by the IEEE electronic engineering association.

The startup said at the time that it had developed a chip capable of generating a motion plan in under a millisecond. That made it hundreds of times faster than competing products on the market, Realtime Robotics claimed. The chip was described as being a field-programmable gate array, a type of processor that companies can optimize for specific use cases to boost performance. 

Alongside its custom silicon, Realtime Robotics has developed algorithms that can create a map of all the possible motion plans with which a robot can carry out a task and then encode the map into the chip at the hardware level. This method reportedly helps reduce processing times. Realtime Robotics’ technology generates the motion plans based on data from lidar units and other sensors installed aboard a system.

The startup says that its technology has multiple applications. In a factory where robot arms operate in proximity to humans, Realtime Robotics’ motion planning chips can help avoid potentially dangerous collisions by detecting when an employee comes too close to a machine. The same collision avoidance features can make it easier to build automated production lines where multiple robots work side by side.

Realtime Robotics is also promising overhead reductions for industrial customers. According to the startup, its chip’s ability to generate a large amount of motion plans and identify the most optimal approach reduces the amount of custom code that plant operators must write to orchestrate robots’ day-to-day work. The startup claims that the task of recalibrating robots for new tasks can be done more easily as well.

Realtime Robotics’ new $31.4 million funding round will help support product rollouts to clients. The startup counts Ford Motor Co. among its early customers. “Having already realized early deployment success, a broad spectrum of customers and partners are working closely with us to refine features and user experiences, readying our technology for rollouts in their engineering, factory and warehouse operations,” said Realtime Robotics Chief Executive Officer Peter Howard. 

Image: Realtime Robotics