Qualcomm has launched its first end-to-end robotics development platform — a set of hardware, software and tools to help developers build the next-generation of industrial and consumer robots — at Mobile World Congress 2019, taking place Feb. 25-28 in Barcelona, Spain.
Dubbed RB3, the platform will feature a heterogeneous computing architecture with support for artificial intelligence as well as additional camera, sensor and video modules. Qualcomm said the platform will support Linux, robot operating system (ROS) and Amazon AWS Robomaker in order to make it accessible to robotics creators and developers. The platform will also support long-term evolution (LTE) cellular connectivity with future upgradability to 5G.
Already incorporated in devices like companion robots from Sony and Anki, multimedia robots like Cerevo Tripon and Keecker, and robot vacuums, Qualcomm said it wants to enable developers to make the next great robot for agriculture, delivery, inspection, manufacturing, logistics, warehousing and more.
RB3 includes a robotics-focused DragonBoard 845c development board based on the Qualcomm SDA/SDM845 system-on-a-chip (SoC) and compliant with 96Boards open hardware specification to support mezzanine expansions. Optional elements to RB3 include a connectivity board, an image camera, 4K video capture, AI-assisted detection and recognition of people and objects, tracking camera for path planning and obstacle avoidance using visual simultaneous localization and mapping (vSLAM), a stereo camera for navigation and a time-of-flight camera for people, gesture and object detection even in low light.
ABI Research forecasts the market for autonomous robots to exceed $88 billion by 2027. Qualcomm’s RB3 comes at a time when commercial robots need to be autonomous, agile, intelligent and aware of their environment if they are to meet the new use cases emerging in the industrial, commercial and retail sectors, said Lian Jye Su, principal analyst at ABI Research.
Applications such as last mile delivery, retail assistance, tower inspection, construction and mining need robots with the ability to navigate and with agile manipulate capabilities in real-time. Su said this is where RB3 will come into play with support for numerous sensors that manage, control and schedule robots’ functions.
Su said that while it is in the beginning phases, cellular, specifically 5G, will be the preferred connection method for outdoor commercial robots. As 5G reaches economies of scale, the total cost of ownership is reduced — Qualcomm appears to be betting on this possibility with its RB3 platform. While the platform does not yet support 5G, it lays a foundation for augmenting the self-awareness capabilities in the future, eventually leading to fully autonomous robots, Su said.
“In the future, this will enable better collaboration between robots and humans or other machines in their environments,” Su said. “Humans will be able to control robots from a remote distance in near real time. At the same time, a fleet of robots will be able to exchange information as they are connected to the cloud platform via cellular connectivity, and coordinate among themselves to perform a specific task or duty.”