Wednesday, January 28, 2009

Sparsh: Build your own Multitouch Table!

With price points for multi-touch table devices coming in at around $7,000 to $20,000 US its not like everyone who loves the technology can afford to buy one. In fact, just about everyone who comments on Microsoft Surface for the first time talks about how cool it is and then how its too expensive.

Those of you who are not as crazy as I am (I bought my own Surface despite the price point) may want to consider building your own multi-touch surface. That's exactly what a combined inter-disciplinary team of students from the Govt. Engineering College, Bikaner did. They used the open source Touchlib multi-touch SDK and advice from the NUI Group along with some affordable equipment to build a wonderful multi-touch device programmed in Adobe Flash.

Sparsh (Hindu for "touch") is the name of their project and you can see a really compelling video of it and what they programmed it to do. The team (Anirudh, Ashish, Rahul, Sudanshu, and Swami) built their own surface device out of plywood, Plexiglas off the shelf components and a little secrete sauce of their own making.

Specifically the device, according to emails communications with Anirudh, was built with the following components:


Table's Touching Area:
1m X .568m
Infrared LED Illuminator's rating: 120mw/sr.
Projector enclosed: Mitsubishi XD 490U
IR detector: Modified Microsoft LifeCam VX1000, detects upto 30 Frames/second
Graphics Accelerator: NVidiA Geforce


MS Windows XP/Vista, DirectX 11 Sept. Update, Open GL, Adobe Flash CS3, AS3
Tracker: modified Touchlib v1.0 beta

The most expensive component is probably the projector which retails for between $1,800.00 and $2,500 US brand new (used models can probably found cheaper). The least expensive component was probably their own special formula for a compliant surface which is needed in order for their method of gesture detection they used (FTIR). I won't go in to details but that particular component can be very expensive but our friends from India found a solution that costs only $5.00 US per unit.

The team based much of their design on guidance they obtained from the NUI Group. NUI Group is commercial company specializing in multi-touch but they also host a thriving open source multi-touch community where you can find a plethora of information on building your own multi-touch device. Just check out the videos of all the home made multi-touch devices features on the NUI Group web site. It's inspiring.

The Sparsh team is currently distracted with final exams (remember they are students) and so have temporarily halted work on Sparsh, but not for long. After graduation the team will probably, again according to Anirudh, develop of a much larger multi-touch wall. The members of the team are very interested in becoming deeply involved in multi-touch device development and are looking for summer internship. If I had my druthers I would hire them in a second for an internship. If you are looking to fund a multi-touch start-up, Sparsh will probably give you the most buck for your investment dollars. There located in India, they are students, and they are obviously very productive. A good investment for any venture capitalist or angel investor looking to break into the multi-touch device market.

My hat is off to the Sparsh team as well as all the inventors who have contributed to the NUI Group community. Building my own surface device that uses Flash would be a wonderful experience but given my current building skills (I’m dangerous at best) and my schedule (I’m totally booked right now) its not going to happen anytime soon.


JoshB said...

Speaking from experience, if they want to commercialize their device:
1) DO NOT waste time going after investments or loans. Instead they should get jobs and do the start-up on their own time.
2) Think very hard about whether they are going to sell services (writing custom apps for their or other device) or products (selling the devices and common applications). Which business model fits their skills and more importantly, the demand and market
3) Keep the company small small small and don't invite friends to join just because they are friends. Expect to make zero money in the first two years (hence why get other jobs first).

Richard Monson-Haefel said...

Great advice, Josh!

anirudh said...

Great suggestion JoshB, at present we're undergrads in India. our priority at present is gaining some experience first by working with a good research team.
Startup, we want to, but not before 2-3 years from now since writing custom apps etc./algorithmics is what we're upon right now.

Aashish K Mahaswary said...
This comment has been removed by the author.
Aashish K Mahaswary said...

Congratulatins Anirudh...

Richard,You featured a gr8 thing, but one thing which i found is , the official site ( says that these are the students from Govt. Engineering College Bikaner which is affiliated with Rajasthan Technical University...and upto the my knowledge it has no relation with Indian Institute of Technology....

Richard Monson-Haefel said...

I've fixed that error. Thank you!

ecb said...

congrats bos....