You don’t have to watch “I, Robot” or “Star Wars” for the 14th time to get the point that one day robots will likely play important parts in our lives.
After all, robots have been making cars since the 60s, and the early 2000s saw the arrival of robotic dogs and vacuum cleaners. However, despite more than a century of literary robots, they have yet to become ubiquitous in our lives.
One contributor to the slow proliferation of robots to date has been the lack of a standard platform to simplify development of robotics applications.
Tandy Trower, general manager of Microsoft’s robotics group and one of the company’s longest continuous employees with 26 years under his belt, hopes to change that. The company Wednesday released the first community technology preview (CTP) of its new Microsoft Robotics Developer Studio 2008.
The CTP is the first of two planned, with final release of the developers studio targeted for the end of the year, Trower told InternetNews.com. In order to underline the tools’ focus, the name, which used to be Microsoft Robotics Studio, has now been changed to include the word “Developer.”
“It shows that Microsoft has continued to make investments in this area and to focus on improvements that help the developer audience,” he added.
The company shipped Version 1 of the Microsoft Robotics Studio in December 2006, and followed up with a 1.5 release in July 2007.
Besides improvements of 150 to 300 percent in messaging throughput, the new version will feature additions to some existing tools. For instance, both the Visual Programming Language tool and the Visual Simulation Environment have been updated. One new feature in the simulator lets developers record and repeatedly playback simulations. Another is a new floor plan utility that lets developers simulate buildings by inputting their dimensions, while a new feature in the visual programming tool enables advanced developers to decide where code modules should execute.
The Robotics Developers Studio toolset is based around a “general purpose” programming model that enables developers to write “asynchronous” programs. That is not new. “The idea was to create a programming model that would let people write applications that do multiple things at the same time,” Trower added.
As it turns out, however, that is a capability that many different types of applications – not just robotics – could take advantage of. “It lends itself more towards service-oriented programming [where you have] tasks running in parallel that have to communicate.”
The implication seems clear – this could be an enabling technology for Microsoft’s software-plus-services initiative among other possibilities. Another way that the technology might be used would be for programming multiple cores on multi-core CPUs.
For that reason, some of the technology will eventually end up in Microsoft’s more mainstream developer offerings, including Visual Studio, Trower said.
Despite 200,000 downloads of previous versions of the developers kit and between 50 to 60 vendors building products based on it to date – Trower isn’t fooling himself. There’s plenty of hard work ahead. He is optimistic, however, projecting significant expansion in the market for robotics within three to five years.
“That seems plausible,” said Greg DiMichillie, lead analyst for application platforms at researcher Directions on Microsoft. “Never underestimate what can happen in a five year timeframe.”
For one thing, by then there could be a lot of devices in use that are not necessarily perceived as robots – like the Roomba robot vacuum cleaner, DiMichillie said. For another, the establishment of a common platform – from Microsoft or someone else – is likely to simplify creating robotic applications, and thus make them less expensive and, hopefully, more useful.
“If Microsoft can succeed in creating a standard platform, it could make it easier to do,” DiMichillie said. “Microsoft can afford to dabble in things like this that might, or might not, pay off.”
