From tiny toys on supermarket shelves to building-sized material haulers, today’s robots come in all shapes and sizes. And thanks to a range of advancements in their components and technologies, they are also becoming more capable and cost effective.
Robots may be the ultimate intelligent edge device. A robot needs to observe the world using many sensors, and reason about what it has observed in order to develop a plan of action. It then needs to perform those actions quickly and safely, often with limited internet connectivity.
One of the most popular frameworks for building that complex functionality is the Robot Operating System (ROS) maintained by Open Robotics, a mature, open source robotics framework used worldwide for commercial and research applications. ROS’ interoperability, body of samples, and community make it valuable for building an automated solution.
Last fall at the ROSCon 2018 conference in Madrid, we announced an experimental release of ROS for Windows. Since then, we’ve been working with Open Robotics to build out support for ROS. This week at the Microsoft Build conference in Seattle, we are pleased to announce the culmination of those efforts: ROS is now generally available on Windows 10.
Windows 10 IoT Enterprise provides the full power of Windows 10, packaged to meet the needs of IoT and intelligent edge devices. It shares all the benefits of the worldwide Windows ecosystem—a rich device platform, world-class developer tools, integrated security, long-term support and a global partner network.
“We’re excited to add Windows IoT as a supported platform for ROS. The ROS developer community can now take advantage of a wide array features in Windows IoT, including hardware-accelerated machine learning, computer vision and cloud capabilities such as Azure Cognitive Services. I look forward to seeing the next generation of Windows IoT-supported ROS applications.” — Brian Gerkey, CEO of Open Robotics.
With support for ROS, the Windows platform now provides a fast, safe, smart and manageable foundation for robotics solutions that also allows developers to do more at the edge using machine learning capabilities and all the scalability and power of Azure IoT:
- Microsoft Azure Cognitive Services provides AI solutions that can infuse robots with intelligent algorithms to see, hear, speak, understand and interpret their environments using natural methods of communication.
- The Microsoft ROS Node for Azure IoT Hub allows a system administrator to monitor the health of a robot and its tasks by monitoring specific message streams.
- The Microsoft AI platform can act as the brain of the robot, with inferencing capabilities that work across any hardware platform. Using the industry standard ONNX model format trained locally or in the cloud, developers can accelerate machine learning at the edge—meaning the robot can run the models itself without consuming expensive bandwidth transmitting images to the cloud.
These capabilities add to the thousands of behaviors, skills and drivers already developed by the ROS community that can be composed to create the mind of a robot. With the core of ROS enabled on Windows 10 IoT Enterprise, many of these components can be made available to Windows with minimal effort.
Additionally, Microsoft will soon be adding functionality to a community-supported Visual Studio Code extension—adding support for Windows, debugging and visualization to enable easier development for ROS solutions.
I’m hosting a session at Build on Wednesday along with Principal Program Manager Lead James Coliz, so be sure to stop by if you want to learn more and see some of these technologies in action. To get started with ROS on Windows now, please visit http://aka.ms/ros.
And to learn more about what Microsoft is doing at Build this week when it comes to IoT, see Sam George’s roundup of all the goings-on.