Ambient Intelligence for Better Buildings

A short introduction to a sensor network research project at MERL: http://www.merl.com/projects/ulrs/
Data from hundreds of sensors all over MERL ffeed work on better sensor hardware, better perceptual tools, better visualizations, and ultimately a better understanding of how people use buildings and how those buildings can help the people within them. Work completed 2006, video edited 2007. Christopher R. Wren, Yuri A. Ivanov, Darren Leigh, Jonathan Westhues, John Barnwell, Alexander Sorokin, Ishwinder Kaur, and Emmanuel Munguia Tapia.

Category:

Science & Technology

Tags:

• sensor_network
• research
• science
• perception
• ubiquitous
• visualization
• data_mining
• interface
• gesture
• MERL

[crwren]

People do not wear anything. The sensors detect heat, and so there is nothing in the signal that is identifiable. Even if I know what office is yours, I can only tell that someone came out of that office, not that it was you. That said, combining this system with cameras, or RFID readers, or active badges... this system can make those systems much more capable, just as it can make your elevators and air conditioners more capable. Privacy in ubiquitous computing is a tricky problem.

[MrQuincle]

That is so interesting. It was also not clear for me if and what type of sensors were used. To use infrared and no cameras might just be acceptable from a privacy point of view. In case of evacuation infrared is also better than movement detection. Someone doesn't want to fail to detect a person, just because he/she doesn't move!

Is there follow-up work performed or planned?

[gregconquest]

It's less intrusive than video, yet more pervasive. Contrary to what nakedbox22 implies, there is no statement in this video about whether people are personally identifiable with this. Is everyone wearing some kind of transmitter? Or is this just an infrared or similar type sensor? If the data can be tied back to specific people, then that database can be mined looking for trysts between co-workers, the culprit who keeps squeezing the toothpaste from the middle, etc. No word about privacy here?

[nakedbox22]

mike, watch the video again. Hopefully this time you'll understand.

[mike54721]

This can never be used for security. To track the people, the people need to have the transponder on them. The transponder actually inserts data into the wireless network that doesn't disrupt it but allows special receivers to see the transponder. An unauthorized person would a) not have a transponder and b) ditch the transponder if he did have one.

[ScotlandTom]

I agree with DudeParallel. This has some pretty strong possibilities for application in not only security, but also with analysis of traffic flow and movement patterns that could help engineers and architects better design structures to handle movements of people and equipment.

[nanodust]

well done !

[Hazclan13]

Very interesting, i think it could work in most small and large buildings.

[DudeParallel]

No comments? Shocking Youtube, shocking.

Amazingly simple little interface.

Combined with CCTV in key locations, this could descrease security costs and increase overall security. (if done right that is)

This could also be used for generating movement patterns in rooms to provide insight into popular areas / congestion and possible redesigns of the floorplanning.