A few years ago, Tandy Trower, a long-standing fixture at Microsoft, met with many people in the robotics industry to determine what their greatest needs were. Trower met with robotics researchers both in academia and in the commercial industry. He also considered the needs of hobbyists and students. His research resulted in a wish list, similar to the following:
Easily configure sensors and actuators and be able to run them asynchronously
Start and stop software components dynamically
Monitor the robot interactively and as it is operating
Allow more than one person to access a single robot or allow one person to access multiple robots
Reuse software components across robots
The needs Trower identified were simple and fell into some well-covered robotics territory: they involved the basic requirements of operating and monitoring a robot across multiple platforms. The common problem was that each time someone built a robot, the creator used different hardware to build it and different software to operate it. Also, building high-level robots was expensive, and this often meant researchers had to share access to a single robot.
Trower felt that if he could create a platform that allowed robotics researchers to work with robots in a standard and consistent way, he could open the doors of robotics to more people. He also thought that the creation of such a platform, which resolved many of the day-to-day details associated with robotics, would allow researchers to focus on the harder-to-solve problems.