CSE makes sense Next-gen smart grid tech OneBusAway Open Data Kit Chair's message News SWARMS A CSE reunion in Oakland Datagrams Awards Alumni Achievement Award College Diamond Award Events Accessibility capstone Engineering Discovery Days ACM spring barbeque Where the jobs are Kings screened at SIFF
CSE makes sense
Technology impacts our lives every day: how we conduct business, how we communicate, how we are entertained. CSE faculty and students continuously explore how technology influences and improves the world both locally and globally. Three current areas of research affect the greater community in various ways: helping the individual consumer reduce water and energy consumption, encouraging more people to use public transportation, and exploring how technology can improve the lives of underserved populations in the developing world.
photo by Jon Froehlich
The monitor displays a HydroSense
pressure wave created by using the
kitchen faucet (background).
Each water fixture in the home
generates unique pressure waves,
which travel throughout the
home plumbing system, thus
enabling the single-point sensing
Next-generation smart grid technology
In an effort to contribute to the nation’s goal of reducing our overall energy use, UW Computer Science & Engineering researchers have started applying their expertise in sensing, embedded systems, and human-computer interaction to create solutions to help better inform consumers of their energy use and encourage more sustainable activities. Our students and faculty have begun to garner significant recognition, not just in the research community, but nationally through the successful commercialization and dissemination of their work to the larger community.
Shwetak Patel and his students have been developing next-generation smart grid technology for the home. Reducing the use of energy and water is one way homeowners can do their part to decrease pollution and slow global warming. Monitoring utility demand has not been easy or affordable for the general public or utilities until now.
Milestones/Recognitions: Sensing Technology
August 2009: Patel receives TR-35 award in recognition of his research.
October 2009: Team receives Madrona Prize for most innovative research for sustainability sensing.
March 2010: Team receives two best paper awards and nomination at CHI 2010.
April 2010: Zensi, Patel’s startup company focused on the commercialization of these new sensing solutions, acquired by Belkin International, Inc. Belkin will market and sell these eco-devices by late 2010.
June 2010: Jon Froehlich (CSE PhD student) receives College of Engineering Research Innovator Award.
Patel and his team have developed sensor systems that allow online monitoring of water, electricity, and gas use by recognizing the unique signature given off by every faucet, toilet, appliance, and electronic device in the home. Using this system, residents can track their electricity and water consumption and take advantage of off-peak power rates. That data can also be fed back to the consumers to help them better understand their own consumption.
At the heart of the team’s utility monitoring system is a novel pattern recognition technology that, once calibrated, can identify individual components of a home’s plumbing or electrical system. The beauty of the system is that the homeowner only needs to install one pressure sensor, usually at the outside hose faucet or water heater drain valve. The entire home’s water usage can then be tracked fixture by fixture. The signal recognition software is smart enough to distinguish between two or more simultaneous events, such as two different faucets being turned on at the same time.
Similar technology has been developed to monitor electrical systems. As with the water sensor, a single sensor plugged into a conventional wall outlet will detect a variety of electrical events throughout the home, each of which has its own unique electrical noise signal. Machine learning techniques allow the monitoring system to distinguish between each light switch, home appliance, or electronic device.
Screen shot of OneBusAway
Improving the usability of public transit
For tens of thousands of Seattle residents, public transit is the primary way to get from place to place on a daily basis. And, as anyone who has ever taken public transportation knows: buses don’t always run on time. Waiting for the bus on a cold and rainy Seattle day can be a frustrating experience. Fed up with the existing transit tools, Brian Ferris, a CSE grad student, created OneBusAway, a set of tools to make using public transportation easier in the Puget Sound region. Kari Watkins, a grad student in Civil & Environmental Engineering, quickly joined. With the support of Professor Alan Borning, the team has been growing OneBusAway ever since.
Available on the web at onebusaway.org, OneBusAway provides easy access to real-time arrival information for public transit vehicles across a variety of interfaces: web, touchtone phone, SMS, iPhone, Android, mobile-web, and other mobile devices. Participating Puget Sound transit agencies include King County Metro, Pierce Transit, Sound Transit, and Washington State Ferries. OneBusAway focuses on answering the question, “Where is my bus?”
Here’s how it works: A rider using the OneBusAway iPhone app can quickly pull up a list of nearby transit stops, select one, and see a list of arrival time estimates for that stop. If a bus is running five or ten minutes late, the rider can use that time to do something else, like grab a cup of coffee or find someplace more comfortable to wait on a rainy day.
OneBusAway is having a positive effect on transit usage in Seattle and beyond. A large survey of OneBusAway users found that, since using OneBusAway, they were: overall more satisfied with public transit; took transit more frequently; and spent less time waiting for busses. A significant set of users reported feeling safer when using the OneBusAway, especially when waiting for the bus late at night. Additionally, users also reported walking more than prior to using OneBusAway, which gave them more information about how to get home, including walking to a different bus stop, or getting a little exercise while waiting for the bus. As one user puts it, "I no longer sit with pitted stomach wondering where is the bus. It’s less stressful simply knowing it’s nine minutes away, or whatever the case." These survey results were presented at CHI 2010, where it received a nomination for "best paper."
OneBusAway has also been a big hit in the community, with over 20,000 riders accessing information from OneBusAway each week. This popularity has led to several recognitions. OneBusAway was a winner for Best Use of Technology in the Government, Non-profit or Educational Sector in the Washington Technology Industry Association 2010 Industry Achievement Awards. OneBusAway has also been nominated for best nonprofit startup in the 2010 Seattle 2.0 Awards.
Ferris and Watkins plan to continue improving and adding new features to OneBusAway, including adding more regional transit agencies. Since OneBusAway is open-source software, it’s easy for the community to get involved with improving OneBusAway and expanding it to other areas. Ferris and Watkins hope that some day using public transit will be easier for riders everywhere, not just here in Seattle.
Open Data Kit: data collection from mobile devices
UW computer scientists are turning
cell phones into flexible tools for
collecting data in the field.
Here, CSE PhD student Yaw Anokwa (right)
watches an employee of a Ugandan nonprofit
try out the tool he developed.
In industrialized countries, data is relatively easy to collect. In low-income regions, the lack of reliable infrastructure, ubiquitous connectivity, and adequate expertise makes data collection difficult. Current practice is paper-based forms. However, physical collection tends to be slow and expensive, data transcription errors are plentiful, there is no access to relevant historical data, and lag time to obtain usable information is significant.
The exponential growth of cell phone usage and infrastructure in low-income regions has sparked great excitement for using mobile devices to address current gaps in data gathering. In addition to the variety of data — text, photos, location, audio, video, barcode scans — that can be gathered, mobile devices have proven to be dramatically faster at both collecting the data and making it available to decision makers.
Open Data Kit (ODK) has been developed to help fill this gap. The software builds upon the Android operating system, the first comprehensive open-source platform for mobile devices; the ODK suite of tools enables users to collect their own rich data. ODK is designed to let users own, visualize, and share data without the difficulties of setting up and maintaining servers. The tools are easy to use, deploy, and scale. They also go beyond open source: they’re based on open standards and supported by a larger community. The ODK team is working with NGOs in many countries to apply the tools and demonstrate the difference they can make in efficiency and effectiveness in a variety of contexts including public health, human rights, and environmental monitoring.
ODK’s goals are to: (1) make tools modular and customizable so that they can be easily composed into appropriate arrangements for each deployment; (2) exploit open interfaces and standards so that solutions are not “siloed” into monolithic enterprise-level packages that are difficult to understand and maintain; and (3) establish data collection tools at the cutting edge of technology to avoid early obsolescence and make it easier to attract talented developers. There have been nearly 2,700 downloads of the free software, and visitors from over 140 countries have visited the Open Data Kit Web site (opendatakit.org).
The ODK research group is building new layers on top of the data collection application base. These include: using the phones as sensors to turn a single mobile device into a Swiss army knife of instrumentation and to better connect data gathering with historical databases (such as medical record systems). With these applications, the phone can help untrained users with diagnostic and triage protocols. Several courses are being connected to the ODK effort at both the undergraduate and graduate levels; a capstone course and a PMP course are planned for next year so more students can apply their skills to create new features that will make ODK more useful.
ODK is possible thanks to generous support from Google.org and Google Research. In February, Gaetano Borriello, the Jerre D. Noe Professor of Computer Science & Engineering, received an inaugural Google Focused Research Award for the work his team is conducting on mobile data collection for public health and environmental monitoring. ODK also received the Pizzigati Prize, the nation’s top public interest computing honor. Our fall 2010 issue of MSB will present a more in-depth look at current research projects using ODK.