Lessons Learned Building TeamTrak: An Urban/Outdoor Mobile Testbed

Wang, Haodong; Li, Qun

doi:10.1109/wasa.2007.35

Cited by 10 publications

(2 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The TeamTrak mobile computing platform [4], shown in Figure 1, is an array of tablet computers augmented with sensors mounted on a helmet. Each node collects sensor data, shares it via an ad-hoc wireless network, and displays the state of the entire network back to the user.…”

Section: Index Terms -Mesh Networking Mobile Computing Research/edumentioning

confidence: 99%

Work in progress - integrating undergraduate research and education with the TeamTrak mobile computing system

Hemmes

Thain

Poellabauer

2007

2007 37th Annual Frontiers in Education Conference - Global Engineering: Knowledge Without Borders, Opportunities Without Passp

Self Cite

View full text Add to dashboard Cite

To better prepare undergraduate students for graduate school and careers as computer scientists, we have undertaken an effort to increase both expertise and interest in research through a novel course format. The course weaves an active research project in mobile computing with a senior class on distributed systems. To facilitate the instruction, we use a mobile computing platform, TeamTrak, which consists of tablet computers that collect and display information from sensors such as GPS receivers and share data via an ad-hoc network. Students participate in outdoor field tests as a group and individually design and execute their own measurement experiments and engineering projects.We have begun an effort to increase the level of interest in research and an understanding of and appreciation for the scientific method by integrating a research project in mobile computing with a senior-level class in distributed systems. By participating in an ongoing research effort, students not only gain experience formulating hypotheses and designing and executing experiments, but in the classroom setting we can take advantage of any teachable moments that arise and explain why students observed the results they obtained.The TeamTrak mobile computing platform [4], shown in Figure 1, is an array of tablet computers augmented with sensors mounted on a helmet. Each node collects sensor data, shares it via an ad-hoc wireless network, and displays the state of the entire network back to the user. Students deploy the system by spreading out across campus, each student carrying one node and its sensors. Each device keeps a second-bysecond log of its location, sensor state, and peer data, as well as any specific data of interest for a particular exercise, so that sessions can be archived, re-constructed, and evaluated at a later date, as shown in Figure 2.We have scheduled a series of outdoor lab exercises using TeamTrak over the course of a semester. Each lab consists of a briefing on the overall problem of interest, the hypothesis to be tested, the procedures to be followed, and the data to be collected. A sample exercise is to arrange the nodes in various configurations, then observe how the routing protocol stabilizes over time. Students initially carry out predesigned exercises but proceed to later design and execute their own.With this format, we intend to produce these outcomes:• Improved understanding of distributed algorithms such as routing protocols by access to a tangible example. • Improved understanding of the scientific method through experiments where the "correct" results are unknown. • Experience in proposing, conducting, and evaluating experiments and engineering projects in small groups. • Benefit to the research project by exposing the system to real users and unexpected situations. A pilot of this course format is taking place in spring 2007. Students complete a series of instructor-led exercises. They also independently design and execute two projects using the TeamTrak equipment: one a quantitative measurement of ...

show abstract

Section: Index Terms -Mesh Networking Mobile Computing Research/edumentioning

confidence: 99%

Work in progress - integrating undergraduate research and education with the TeamTrak mobile computing system

Hemmes

Thain

Poellabauer

2007

2007 37th Annual Frontiers in Education Conference - Global Engineering: Knowledge Without Borders, Opportunities Without Passp

Self Cite

View full text Add to dashboard Cite

show abstract

“…In our experiences with the TeamTrak mobile testbed [4], an approximation of wearable military computing systems for individual soldiers [11], we have observed that even with sensor hardware of the identical manufacturer and model, error frequently occurs quite independently between devices. GPS receivers placed in the same physical location at the same time and loaded with identical almanacs will not necessarily acquire the same group of satellites with the same geometry, or may acquire different numbers of satellites, resulting in varying degrees of accuracy.…”

Section: Introductionmentioning

confidence: 99%

Cooperative Localization in GPS-Limited Urban Environments

2010

Self Cite

View full text Add to dashboard Cite

Abstract. Existing localization techniques such as GPS have fundamental limitations which preclude deployment in urban canyons or areas with inconsistent network availability. Augmenting GPS requires specialized infrastructure or tedious calibration tasks which limit general purpose applications. In this paper, we examine the utility of cooperatively sharing location data among connected nodes in order to correct positions with high measurement error in GPS-limited environments. Using simple data sharing and filtering techniques, collaborating users can substantially reduce overall localization error in dead reckoning systems where nodes may have a broad spectrum of location quality. We examine system parameters necessary to fully exploit cooperative localization based on empirical error models and show that mean position error can be reduced by up to 50 percent for given application scenarios. If distance measurement is available, filtering location information based on estimated error and confidence can improve accuracy of pedestrian dead reckoning techniques to approximately that of GPS using trilateration.

show abstract