Wednesday, February 01, 2006

NXT Gaming

Yesterday, I wrote about the desirability of an NXT marketplace, from a commercial point of view. Today, I’d like to come back to that point, but looking at it from an educational point of view.

Lego is replacing the old Lego “dumb” motors in the Mindstorms RIS set with new Lego “smart” motors in the Mindstorms NXT set. The smart motors encode their rotation to an accuracy of 1 degree, which allows the NXT controller to move the motor to a specific angle, or count the amount of rotation. At least, that’s how I currently understand the motors to work, based on the bits of information available.

This change to smart motors is a big deal. Not because more complex robots can be built with it: electronic engineers could make motors do smart things before. But because complex robots can be built easily. With NXT, having a robot wave its arm or turn its head back and forth is child’s play.

I believe the same thing should happen when it comes to NXT software. As part of Mindstorms NXT, Lego will be providing a new development environment based on LabView. This new environment makes it easy to create new software. But it can only go so far.

For instance, every driving robot can benefit from the concept of “deadreckoning”. Deadreckoning essentially determines the position of the robot by counting wheel rotations. However, the formulas of deadreckoning are certainly not child’s play. They have to take into account wheel slippage in curves, and involve lots of trigonometry. On the other hand, a twelve-year-old could do deadreckoning easily if she had a pre-built component “deadreckoning” in her programming environment. The component is essentially talking the rotational sensor, and adding computed functionality to it.

Just about every sensor can be expanded to deliver a lot more information by using advanced mathematics and science. A child playing with the Lego set will be stimulated in his creativity and his exploration by using these libraries.

But the concept only gets started with basic sensor libraries. Building advanced robot control systems is more than putting basic capabilities together. Architecture of advanced robot intelligence can be daunting, and is certainly beyond the level of a twelve year old. Providing ready-made advanced robot games will teach the child a lot about what more can be done. Kids like to imitate, and they will certainly start to explore how a robot manages to achieve particular abilities. Plus, a good robot game should allow the kid to customize it, and add to it. It should be fun to see how kids tweak advanced Lego robots to make them better or more personal.

NXT Games are really just building on above concept. By selling the NXT Games, more time and effort can be put into creating them. Hopefully, this results in astonishing levels of competency and interactivity for NXT robots. This, I believe, will result in improved educational value for children, as well as in improved revenue for Lego.


At February 02, 2006 12:36 AM , Blogger Jim Kelly said...


What would be REALLY nice is for LEGO to introduce a new sensor every 3 to 6 months at a reasonable price ($15-20). With the sensor would be a CD-ROM that would introduce the functions of the new sensor. It would also contain APIs or more complex 'modules' that would work with the LabView software. And then finally, the CD would also contain instructions for 1 or more robots that are built with the sensor and suggestions for further play with it. I'm a big believer that the user (kid or adult) should be learning something about electronics or physics or programming in a simple, easy-to-read format with plenty of pictures and/or animations.


Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home