CITRIS 2006 Symposium
By mkupfer on Dec 21, 2006
The UC Center for Information Technology Research in the Interest of Society (CITRIS) had a half-day symposium on December 14th, with a panel discussion, a couple talks, and a poster session. The Cal alumni magazine had done an article on some of the CITRIS work not too long ago; I wanted to learn more, so I signed up.
I'm glad I went. Some interesting points were made during the panel discussion and talks, and some of the student projects were fascinating.
During the panel discussion, Beth Burnside (Vice Chancellor for Research) pointed out that universities can look at technology transfer offices in a couple different ways. One way to look at them is as an income source for the university, e.g., via patent licenses. But it's not a given that the office will actually produce any income worth noting. Another way to look at technology transfer is that it's a way to get research results applied in the outside world. That is, it's a way for the university to make a real difference in people's lives.
One of the talks about energy and global warming. Paul Wright (CITRIS Chief Scientist) talked about things like energy conservation and energy sources other than fossil fuels. It turns out that British Petroleum is soliciting proposals for a university research center into biofuels. I was pleased to hear that BP is doing this. If companies think of themselves as fuel or energy companies, not oil companies, they and we will find the transition away from fossil fuels a lot easier.
After another talk, someone from the audience asked if the speaker had any thoughts on how to deal with the combination of too much information and illiteracy. If you look at the typical results from a Google query, there are pages and pages of hits. People who can read can skim over them to find the ones that are most likely to be interesting. What do people do who can't read? The speaker answered that there is ongoing research into voice recognition and artificial speech, but that misses the point: how does someone quickly pick out the interesting results from a long list that is being read aloud, one entry at a time?
I was most interested in Eric Brewer's talk about technology and infrastructure for emerging regions. Some of what he talked about was how to be successful when working in developing regions. For example, they spent a fair amount of time up-front talking with various non-governmental organizations (NGOs). Some of the organizations were actively looking at how to use new technology, others weren't really interested in new technology. Since new technology was what CITRIS has to offer, it made sense to partner with NGOs that were clearly interested in that area.
Another thing they did to be successful was to do small deployments every 6 months. There was some local skepticism about CITRIS due to past interactions with other groups that had made big promises but didn't actually produce any results. These smaller but frequent deliveries helped counter that skepticism and build trust.
Eric also talked about more technical issues, and I was able to learn more from talking to students at the poster session. One of the projects is a telemedicine project in India. Eric said that 70% of the blindess in India is treatable, but only 7% of the rural patients are able to get to a clinic for treatment. Even if the clinic is within walking distance, there's no guarantee you'll be seen when you get there, so many people don't bother. Telemedicine is an obvious remedy to this problem, but bandwidth is a challenge. Satellite links are too expensive, and wireless bandwidth drops off sharply when there are long distances between links.
The bandwidth problem is due to the protocols that are normally used, which rely on collision detection (like ethernet). As the link distance increases, the propagation delays lead to more and more collisions, which kills throughput. So they implemented a synchronous (time-sliced) protocol, which doesn't rely on collision detection. The change in low-level protocol is entirely transparent to higher-level protocols.
And it turns out, they don't need fancy telemedicine facilities to be effective. Just having a video conference link means that patients can get an initial screening. The ones who need more extensive examination or treatment can then set up an appointment at the central clinic.
I'm very pleased that the University of California is doing this sort of work, and I'm looking forward to hearing more about it in the future.