October 14, 2004 Minutes

John introduces Mark Younger of Duke Engineering Living Technology, who will talk about the SmartHouse project. The university is giving students a large budget to build a large house to live in called SmartHouse. Building begins in January.

Mark says he has been working on the project for the last year. SmartHouse will be behind the Freeman Center , off Campus Dr. and Swift. Ten students will live there each year, as long as they are researching a system in the house. The research scope is defined as the 3 E's: Energy and Efficiency, Environment and Health, and Entertainment and Communications. We see research moving towards environment/technology interests. The goal is to give undergrads more practical design experience, and also to make innovations available to homeowners.

The design of the house is half-done, and groundbreaking is planned for March 1, 2005. Key design issues we are dealing with right now are: flexibility/adaptability (probably the #1 priority); integration; communication (devices talk to each other, maintenance reports from appliances); monitoring/reporting; and performance. The project now has 40 students doing research in different teams to create the house.

John Board asks where the faculty oversight is.

Mark says it is within the structure of the teams.

John mentions that faculty members are purposefully on the edges of the project.

Mark says the timeline has the architects hired in March 2005, 9 months of construction, and students moved in spring semester 2006.

They are working on planning the network right now, and dealing with centralized vs. decentralized control and non-proprietary open systems. It doesn't seem like anyone in industry has come up with a residential solution for the kind of networks we want yet. Some larger companies are working on systems for commercial buildings, but they may be high-cost and not work for a residential house. There are also smaller-scale systems like MisterHouse. We are trying to find a balance between these two to create a backbone that can allow students to research and homeowners to see future technology. We're looking at the issue of having an enterprise system vs. research network. We also need standard controllers.

SmartHouse will be a good test-bed for the university. It is a possible test-ground for OIT devices like VOIP.

Some possibilities they are looking at for the house are a media-on-demand system; piezoelectric tiles (material uses wasted energy from foot to use for power); acoustic monitoring; a sensor platform; an artificial intelligence temperature system; and LED lighting to be used for color and white light.

Additional projects include an Al thermostat, occupant tracking, Doppler radar for perimeter security, fire reporting, and noise cancellation.

Ken Hurst asks about documenting/filming the house. He can see this turning up on HGTV or another TV program.

Mark says there are legal constraints on filming, but they will be open to media interest.

George Oberlander asks how much flexibility the house will have to strip out and do things again.

Mark says they are planning to make it as flexible as possible. Flexibility is the number one priority now. They believe the key is making a backbone around which other systems can be built and rebuilt. He adds that they have designed the house so can you change walls without taking down the foundation or other permanent structures (‘smart walls').

George adds you might want to adjust indoor lighting levels to test seasonal effective disorder.

Mark says that is a possibility.

Ken Hirsh says that there is a lack of variety in the age of house residents, so they might take that into consideration.

Mark agrees that there is a lack of age diversity, but they are trying to get other people involved. They would also like to get people from other schools within Duke to participate.

Mike Pickett says it might be worthwhile to talk to people in the health system in preventative medicine to figure out if they can research systems to determine if people are about to become sick, i.e., things in the house that could automatically sense and communicate to residents.

Scott Smith says they working on a device residents can wear on belt to monitor the body for major problems like arrhythmia.

Mark adds that the device can automatically send notification of problems to a doctor or other medical professional.

Molly Tamarkin says there is an interesting tension between what is efficient and what is experimental, and asks if they are leaning towards one principle more than the other.

Marks asks by efficient do you mean of what is established?

Molly says she is asking how they will make a decision about something unknown and experimental vs. something known to work pretty well.

Mark says they acknowledge that projects won't pay for themselves. By limiting the budget students will work with, it will focus their projects. But they are favoring the experimental.

Brian Eder asks if they are planning on using different test groups like families.

Marks says there are no definite plans now, because there is not enough room the house; maybe in the summer when students are gone.

Brian asks what about sitting down and talking to families.

Mark says community outreach is definitely a part of the plan, such as an informational Web site.

Mark asks if there is an opportunity for collaboration between the SmartHouse group and members of ITAC.

John says letting this group know about it is a good first step.

Tracy asks how the project has been looking at benchmarks or other groups doing similar projects.

Mark says the University of Pennsylvania has similar house.

Tracy says she just wanted to make sure they were looking at others similar projects.

Mark says the different between the Duke SmartHouse and existing similar projects is no one else has students living in their houses.

Mike says right now there are 2 main PBX systems, and 6 others supporting 49 departments. 5 of these 6 are no longer supported by the manufacturer, so if something goes wrong, we may or may not be able to fix them. They are trying to evaluate the magnitude of moving away from these 6 systems and using one of the 2 main systems. Because of industry advances, they are now hopeful that the functionality needed for the departments supported by the 6 smaller systems can be handled by the primaries.

Tracy asks if there are more really than 6 systems, if these 6 represent many smaller systems?

Angel says there are 49 systems divided into 6 types.

Mike says they looked at how can they fast track into this migration. They are looking at a 1 yr plan that will start around January, and have looked at a budget. This includes updating telephone units, labor, installation, etc. Main point: if these break, we have no relief. We may or may not be able to get parts to fix them. At the same time we don't want to disrupt services. We will be in touch with departments to make sure services can be replicated on the new systems. If major issues arise, it will probably affect the overall schedule. The goal is to provide departments with equal or better service, and make them aware of new features.

Angel says another benefit is that this in combination with an RFI for a voicemail system will allow anyone with service through Duke to share messages. We are trying to get everyone on single voicemail platform.

John Board says he is glad to hear this is happening. The phones at least will work; will there be other operational savings?

Mike says other savings haven't been completely calculated; as old systems go out and need less support, they may be able to reallocate staff. They also want to have adequate time before several long-time employees retire to transfer knowledge to newer employees.

Robert asks if there will be an opportunity for an experiment of the migration with a smaller group.

Mike says yes, we have small community that would be easy to test. It is definitely a point to consider.

John Board asks if there is a short version.

ike says current 2 primary systems can support the other departments. It may not be a long-term solution, but at least everything will be on the same platform. It is a case of consolidation on a single point so future migration will be easier.

Mike Pickett says CSG used to be called Stone Soup; it is a group of universities, invitation only, techie types, who attend a one-day workshop on a tech project that is a challenge to all. The workshop then goes into a meeting with a policy discussion. The common theme is that there are common problems and solutions facing our campuses.

The focus for the most recent meeting was messaging and collaboration services. Mike presented on e-mailing at Duke. Brown reported that it is using a lot of Microsoft Exchange, and says it's going very well. They only have one support person for entire university.

There was a short workshop on the impact of pervasive computing. Tracy talked about iPods. Dartmouth did an iPod pilot with 22 iPods. There was much talk about where tablet PCs are going. On average, for universities attending 33% of e-mail passing through systems is junk/spam. The annual cost per useful message for universities ranged from 7 cents to $1.69 per message.

Michael Gettes says the calendar discussion is continuing, and we have opportunity here to fix things. OSAF is leading the effort on CalDAV. Vendors like it. OSAF is now considering development for a CalDAV server, but they would like to have someone else fund it. There is also a group that has been formed to talk about iCalendar. Right now there are aspects of it that are too complex that need to be simplified.

In September 2004 there was a Calendar Roundtable in Montreal , sponsored by Oracle. Attendees included IBM/Lotus Notes, Novell/Groupwise, Novell/Evolution, Yahoo!, Oracle, Apple, Statalabs, OSAF, Mozilla, Nokia Isamet, University of Washington , UC Berkeley, MIT (car trouble), Duke, NASA-JPL (absent). Microsoft was coming but the staff was called back at the last minute. All agreed that ietf-calsify was critically important and that CalDAV was the most interesting way to go. All think now is a good time for calendar interoperability.

Stanford and University of Wisconsin joined calconnect. Duke will join if the vendors join.

Tracy adds it cost a couple dollars to join, a minimal amount.

ike says Duke held up as poster child for problems in calendaring world. “If we solve Duke's problems, we've won.” By end of November we'll know where the vendors stand. Calconnect.org has lots of information, and is trying to start an organization. Note: UMA = unified messaging association.

John Board says yesterday Duke was the victim of a forged e-mail. He asks for an update, and who is leading the issue.

Tracy says OIT staff is leading it. Wednesday night around 1:30am 1,700 messages came through the OIT mail servers to people on campus: faculty, students, and staff. There may have been additional messages, and there is some, but not much, evidence they came through other systems.

Chris says he has started collecting information on forged e-mails that went through other systems.

Tracy says the matter began when they got a few reports of the e-mails in the morning. They quickly realized it was a forgery. Tracy asks Chris to elaborate.

Chris says complaints were received in the morning, and the e-mails looked like forgeries early on. They looked at where the e-mails came from, and identified it as a Verizon DSL connection from Los Angeles . The IP address said it was from L.A. , a trace route was made to L.A. , and Verizon said it was from L.A. They tried to determine if it was enough of a criminal act to try to pursue the person. They forwarded the information along to the Duke police department, who said because there were no specific threats of violence we probably couldn't get a court order. Duke police contacted Verizon, and they said they will only give out information in presence of a court order. We now have a denial of authorship of the message from both alleged senders; we have proof it came from a computer in L.A. ; all of this suggests it is a forgery.

Tracy adds that a similar incident occurred in Michigan , and it was also traced to California .

Chris says there is a need to talk about how the e-mail addresses were harvested. He thinks it was done from websites, because one message was sent to ‘webmaster@phi.duke.edu.'

Tracy adds that another reason to think so is only a few messages went to undergrads.

Chris is checking with departments to see how many of the forged e-mails they received.

Robert Wolpert adds that the e-mail addresses didn't have to be a harvesting specifically for this purpose.

Rafael Rodriguez asks if this is a case of identity theft, and if so, why doesn't Duke pursue it as a criminal case?

Chris says he doesn't think it rises to the level of identity theft that could be prosecuted.

Rafael says the state attorney's office contacted him about identity theft in the health system; we might want to contact them.

John says law enforcement doesn't have a definite policy on these things, which makes it difficult to prosecute for this kind of forgery.

Shailesh Chandrasekharan asks how easy it is to forge an e-mail like this.

Chris says someone just has to change the name of the sender in ‘Preferences'.

Owen Astrachan asks if the message in Michigan is likely to be from same person that sent the forged message to Duke.

Chris says Michigan contacted the company and found out the source was a Verizon DSL line in southern California .

Scott Smith asks if there is any device to notify OIT when 1000+ e-mails from one sender appear on the Duke system.

Chris says it's not easy; once there was a UNC project that wanted to contact Duke students. We need to be careful before we implement such a system.

Mike Pickett says it seems like there might be something behind pursuing this as a case of identity theft. Even if Duke can't do so, the two people whose names were used could pursue it.

Lynne O'Brien asks if other people outside Duke got similar messages.

hris says several people within the institution are getting negative messages.

Angel says we are set to move 210 of 280 OIT people to the first and second floors of the Strickland Building , December 10-11 and 17-18. They are now down to details like coat hooks, wastebaskets, parking, etc. We hope to have everything finished before the holidays. At this point we are looking to figure out what to do with the existing furniture. We are going with an open space with few walled-off areas. The courtyard area is done, and has nice landscaping. Colleagues from financial services are moving to the ATC in February. There's lots of security out there.

John Board asks where all the people are coming from.

Angel says they are coming from University Tower , Erwin Mill, the North Building , the group at Kangaroo, the Telcom building, and South Square II. Telephone and computer operators are staying on campus.

John asks where people can go to talk to OIT people, that is, how campus will get contact with OIT.

Angel says the Help Desk is where most face-to-face contact currently occurs. The rest goes over phone/e-mail. Lots of customer service is moving out there.

Kyle asks if there will be a shuttle between campus and the ATC.

Tracy says that's being worked out. There is also public parking around there. Tracy wants to have an ITAC meeting at the ATC so everyone can see it.

John asks Tracy if she will be moving out there.

Tracy says she'll have an office both places.