Microsoft Robotics Studio

[Zak698]

You know microsoft dosent stand a chance against first... because some day we will take over the world.... not the robot of bill gates... but that dose look very intreasting none the less

[Elgin Clock]

I find it funny that the person who is the general manager for Microsoft's Robotics Initiative name is Tandy.

Shouldn't he be working for Radio Shack, and not Microsoft?

[tiffany34990]

It looks really interesting. A lot of companies are trying to achieve FIRST's goal.
One main factor is trying to make kits that are not expensive and easy to use.

[Donut]

I'll wait until they actually show me a kit before I decide if this will be any good. I don't know though, Microsoft and robot kits...?

I like their idea of some more standardized robot systems. It'd be nice to easily mix multiple robot kits/pieces together into one "super" robot.

[Madison]

Originally Posted by Donut

I don't know though, Microsoft and robot kits...?

Hey, we've been doing okay so far

[KenWittlief]

I dont see any mention of kits in the news release. They are only taking about an operating system, not hardware.

Personally I cant imagine any real time system with mechanical components running on Microsoft based SW. How does the old joke go? If Microsoft built cars then.... ?

[petek]

Originally Posted by KenWittlief

I dont see any mention of kits in the news release. They are only taking about an operating system, not hardware.

Personally I cant imagine any real time system with mechanical components running on Microsoft based SW. How does the old joke go? If Microsoft built cars then.... ?

When it was introduced in 1991, I wonder how many people thought that Visual Basic would become the defacto standard office automation language and that by now supposedly over 50% of all programmers would use it?

According to an article I saw:
...others, not just Microsoft, will be able to offer software development tools for use with the underlying Microsoft Robotics Studio system. And while the system will run on Windows, it can be used to create programs for all types of robots, not just those using the Microsoft operating system, Trower said.

Maybe even ones using Microchip controllers...

[Donut]

That would certainly make code transfers easier.

[Tom Bottiglieri]

This is not a robotics kit. MRS (Microsoft Robotics Studio) is an SDK that allows robot programmers to create and manage code for robots.

Here's a good video showing some of the features of the Microsoft Robotics SDK.

http://channel9.msdn.com/ShowPost.aspx?PostID=206574

[Astronouth7303]

While a universal platform is a good goal, isn't the differences between controllers where their power comes from?

I'm watching that hour-long intro.

Within the first 10 minutes:

• It's sounding like a good way to introduce hard-to-find bugs: deadlocks, race conditions, and any other concurrent bugs
• Managed code = .NET; if it has the performance of .NET 2.0, I'm not touching this.
• Simulator = lots of work while every team creates a realistic, accurate, high-res model of their robot
• JScript, VBScript = I'm thinking running Microsoft OS (likely based on NT) on your robot

Did we mention that our current processor runs only at 20MHz? (Compared to the current 4000 MHz.)

Of course, it could be that it compiles to .NET (CLR) and converts that to the local machine language. (Much like JIT compiling.) Maybe this is an effort by Mircrosoft to show the versatility of .NET.

Nice lawnmower, though. (The coffee maker is a LIDAR unit, BTW.)

I'm willing to bet that this is going to be about as useful to us as CoreChart (that graphical assembler).

EDIT: At about 16:03, notice the Windows Live logo.

[Astronouth7303]

So I finished that intro. Here's what I noticed.

• Windows logo on NXT
• IE required (no Gecko goodness)
• Jumping boxes in simulator
• Requirements of a PC to run this
• This does fulfil some of my goals for FURC
• The 1 female shows up in the last 8 minutes and is only around for 4 (in an 58 minute introduction)
• Not a single CMUCam, but plenty of LIDAR boxes. (And guess which is cheaper)
• Can handle many platforms of robots, but only 1 platform of PC
• .NET through and through
• A PC required???

"2 lines" to express lots of complex information. Sounds like a good way to overload one's brain.

How many times will you hear "Oh. I guess the simulation was wrong."?

Much like CoreChart, this has potential but the jury's still out.

[Andrew Rudolph]

Microsoft has set up a blog for the devolpment of the program,

http://blogs.msdn.com/microsoft_robo...o/default.aspx

[Adam Y.]

Originally Posted by Astronouth7303

While a universal platform is a good goal, isn't the differences between controllers where their power comes from?

I'm watching that hour-long intro.

Within the first 10 minutes:

• It's sounding like a good way to introduce hard-to-find bugs: deadlocks, race conditions, and any other concurrent bugs
• Managed code = .NET; if it has the performance of .NET 2.0, I'm not touching this.
• Simulator = lots of work while every team creates a realistic, accurate, high-res model of their robot
• JScript, VBScript = I'm thinking running Microsoft OS (likely based on NT) on your robot

Did we mention that our current processor runs only at 20MHz? (Compared to the current 4000 MHz.)

Of course, it could be that it compiles to .NET (CLR) and converts that to the local machine language. (Much like JIT compiling.) Maybe this is an effort by Mircrosoft to show the versatility of .NET.

Nice lawnmower, though. (The coffee maker is a LIDAR unit, BTW.)

I'm willing to bet that this is going to be about as useful to us as CoreChart (that graphical assembler).

EDIT: At about 16:03, notice the Windows Live logo.

Well that shouldn't be a problem considering the fact that Parallax is on that list. I wonder how extendable it is.

[Astronouth7303]

Originally Posted by Adam Y.

Well that shouldn't be a problem considering the fact that Parallax is on that list. I wonder how extendable it is.

Parallax hasn't been used since 2003. InnovationFIRST released the Microchip PIC18F-based controller in 2004.

[Tom Bottiglieri]

If I remember correctly, this platform brings the idea of a PC base station into play. The robot has standard driver software running on its local hardware, and is connected to a PC (via LAN, Bluetooth, Serial, etc..) An app running on the PC reads and writes data to the robot based on user input and pre coded "stuff". The robot speaks the same language as the app and knows how to move accordingly.

I'm not sure how far the robot feedback is abstracted. If it gets processed on the PC, we would be able to do some pretty neat stuff with this. I'm definitely going to give it a try.