Compared to This, the Market for Cell Phones Looks Extremely Limited

By Al Riske

21.Feb.08 - Sometimes the smallest things can make a big difference.

Take mobile phones. They're tiny but drive huge amounts of traffic in telecommunications (where Sun sells an awful lot of its systems).

But, as Roger Meike puts it, "There are 6 or 7 billion people in the world and we've already got Java on billions of cell phones, so at some point it's a limited market."

A few years ago Meike, a senior research director in Sun Labs, began to ponder an interesting question: What would be bigger than cell phones?

The answer was something much smaller.

What got the high-tech industry excited about mobile phones was how quickly they started to outsell PCs.

"So what is going to be bigger than that?" Meike says. "What will be something where instead of having one you have hundreds or thousands of them?"

Imagine a network of small programmable objects. A network of things, if you will. Things that can detect changes in light, heat, humidity, and motion. Things that can form ad-hoc wireless networks. Things that can make good things happen.

"These little devices may be something where you have hundreds of them in your car and thousands of them in your home and office and they're just surrounding you every day," Meike says. "The opportunity there is huge."

That was the motivation behind the Sun Small Programmable Object Technology (Sun SPOT) project that Meike leads.

Project Sun SPOT is a hardware and software platform designed to overcome the challenges that currently inhibit the development of the emerging network of things.

"Our first step was playing around with the existing sensors that were out there and we didn't find anything that really fit our needs. It was very difficult to get to the level you needed to start programming and the barrier to entry for a programmer was very high -- there were a lot of special skills you needed to develop and special knowledge you had to have about these particular devices as opposed to those," Meike recalls.

If he and his team were going to kick-start a revolution they were going to have to invent something much simpler.

"Hopefully that would speed up the adoption of these types of devices and speed up people inventing the next cool thing," he says.

What the team came up with is a small wireless transducer running a Java virtual machine that acts as both operating system and software application platform. Each one has a processor, some memory, a radio for communications, and various inputs. Most include a light sensor, a temperature sensor, and a 3-D accelerometer.

"Normally this sort of stuff is done by embedded system programmers and they are a hearty lot. They have their bit tweezers and they are right down there working on things, bit by bit. Sometimes that works out great but it's such a small number of people who have that skill set that you're really limiting who can be inventing new things with this technology," Meike says.

"We have created everything from device drivers and development tools to demos and tutorials so users can quickly create their own embedded wireless applications. There are so many many many more people who know Java that we're getting this into the hands of whole different sets of people."

The Sun SPOT team wants to inspire others -- and be inspired by them.

"We've had great success with Sun SPOTs even in art schools where people have built amazing things even though they're not really programmers," Meike says. "That's not what they want to do with their lives but they have something else to say and they can use Sun SPOTs to say that."

Consider what happened at Art Center College of Design in Pasadena, California, where the school hired a Java programmer to assist students in creating functional artworks.

"They came up with all sorts of things that were really amazing, including new types of musical interface devices that, as you shake them, would create music," Meike says. "Dancers would tie them to their bodies and dance around and make music."

Two students designed a flock of helium-filled blimps controlled by Sun SPOTs.

"These blimps would fly around in this big space and flock together and 'eat' together off this other Sun SPOT device. It was fantastic stuff. The fun thing for us is not only were these students doing crazy out-there things and being inspired by us but they inspired us," Meike says.

"They were working on swarms of devices and we said, 'Hey, this is an interesting idea: Lots of devices cooperating to get to some goal. How can that be useful in our world?'"

As it turns out, the team quickly found a business application.

"We had a customer come to us with a shrinkage problem. Things would disappear somewhere along the line during shipping -- somewhere along that transportation chain that they didn't own," he says.

"We suggested a system where they would have sensors in their boxes and when they put them in the truck all the boxes would be comparing notes with each other," Meike says.

I'm bouncing. Are you bouncing?

Yes, I'm bouncing.

I stopped bouncing. How about you?

"Then if only Box 45 is bouncing, all the other boxes on the truck, as a swarm, can notice that something peculiar is going on," Meike says.

Which is very different from trying to monitor the shipment through global positioning satellites, or GPS, he notes.

"With GPS, you'd have a hard time tracking that. How would you know if Box 45 just moved three feet off the back of the truck and somebody opened it up? But this is very elegant. You'd know if suddenly there's light inside Box 45."

Another practical application came about because one of Meike's teammates happened to be having lunch with someone on the Project Blackbox team

"The two of them started to realize that Sun SPOTs could be really useful because with the Blackbox you've got this big container filled with expensive computers and when you unplug it and stick it on a truck you've got no way to know what's going on with them," Meike says.

"We decided very quickly that we should put together a little system for them to put in Blackbox. We put in some humidity sensors and used the accelerometers to sense how things shake on the truck and if there was anything we needed to worry about there. We worked with their team to make sure we were getting data that was actually useful to them."

Two weeks later when the first Blackbox went on tour it had about a dozen Sun SPOTs inside measuring various things -- light, temperature, humidity, acceleration -- and storing data on the same kind of flash cards used in digital cameras.

"Then we hooked up a cell phone to it as well so it would be able to call home and we could track it," Meike says. "We knew exactly where the Blackbox was as it traveled around the country, and we were collecting data about what was going on inside it."

Meike continues to be amazed by the applications people are coming up with. They range from monitoring water quality in Malawi, Africa, where cholera outbreaks have been a problem, to collecting data about the Amazon rainforest with sensors at both ground level and high up in the canopy.

"One of the fun things about this project is it's about building a community. We want to inspire people and once they're inspired we want to get some of that back from them -- some of their inspiration back -- and so far it's been really good," Meike says.

"Every time someone comes up with one of these great applications, it really pushes the team to go farther. We're so much farther than we could have gotten if this was just a project we were doing by ourselves in the lab. If we get this out into the world where other people can start doing this stuff, we get the benefits of that -- we get inspired by it."

More futuristic ideas include environments that recognize you and remember your preferences -- a room that adjusts the temperature to your liking, a car you've never driven that knows how you want the seat adjusted.

To get there, though, Meike sees a multi-step process.

"First we have to get people excited about these types of devices and hopefully get them excited about Java and about Sun in the process. In any case, many of these devices will be generating data because they have a lot to report. That would mean data that needs to be stored and processed, commands that need to be sent over the network. All that stuff feeds right into our existing server business. So we want this to happen and it would be really great if it happens with a platform we design, but overall we want to enable the world to build new things because that's good for us."