The 8th ACM/IEEE International Conference on Information Processing in Sensor Networks

April 13-16, 2009, San Francisco, USA

Extreme Sensing 2009:
The Tight Rope Walk

The Eight International Conference on Information Processing in Sensor Networks (IPSN '09) IPSN '09 will be hosting the Extreme Sensing competition April 13, 2009. Each team will need to build a sensor system using 7 sensors or less that can perform dead reckoning as people walk along predefined paths through a 10m x 10m arena. Each team can use up to 5 body sensors and at most 2 sensors placed in the arena.  In Phase 1, all teams will track conference attendees, estimating their final position relative to a known starting position.  Results will be broadcast live during Phase 1, providing each team an opportunity to discern how the other teams' systems work and how they might fail. In Phase 2, this information can be used to prevent the other teams from scoring more points, as each team gets one chance to walk through the arena themselves while suited up with their opponents' sensors.  

Rules of Play

Each team must build a sensing system that can perform dead reckoning as people walk through a 10m x 10m arena, which will be located in the poster/demo session of the conference. The goal is to estimate the final position of the person given the initial position. Teams can use up to 5 body sensors, which may include accelerometers, gyroscopes, magnetometers, etc. Each team can also setup at most 2 sensors in or around the arena, but predefined paths must remain unobstructed for people to walk, as shown in the diagram below.  The predefined paths will not be revealed until the day of the competition, and will be marked with tape. Examples of sensors placed in the arena might include passive infrared motion sensors, active infrared break beams, ultrasound, dopplar radar, weights sensors, etc. These sensors can be useful for on-line calibration of the body sensors.  The sensor systems have three restrictions: 1) no system may have data cables more than 12 inches in length* 2) no system may impede the motion of the person being sensed, and 3) no system may use more than 7 pixels in total, eg. no cameras with more than 7 pixels, and no more than 7 single-pixel** sensors. Signal emitters will be permitted, as long as they do not interfere with other teams' sensors and as long as they do not help the system emulate more than seven pixels (see below**). Each time a person walks a path with a team's sensors, that team will be required to update a server with its current position estimate at least once per second.  The path estimate will be visualized and projected onto a wall.

During Phase 1, between 5-10 volunteers from among the conference attendees will walk along predefined paths in the arena, while being tracked by the various teams.  Each volunteer will walk the same path for all teams, and can wear body sensors from multiple teams simultaneously, if possible.  The volunteer must walk along the path, with each foot landing on the path in front of the other foot, like tight-rope walking -- i.e. the volunteer cannot veer from the path and cannot walk backwards.  Aside from these restrictions, however, the volunteer can move the body in any way that he/she chooses.  The teams must be able to outfit each volunteer with sensors in 1 minute or less, and must put a sign on the volunteer to indicate that he/she is wearing that team's sensors.  Ground truth will be obtained by entering the path number and a set of team numbers into a computer before the volunteer walks the path.  During Phase 1, contestants will not be allowed to enter the arena.

In Phase 2, each team will be given one chance for a team member to walk a path with every team's body sensors. The goal of each team is to score points for themselves without scoring points for competing teams.  The team can choose any of the predefined paths, but must obey the same rules applied to the volunteers in Phase 1.  The team member must wear all teams' sensors simultaneously, to the extent possible.  If the different teams' sensors cannot be worn simultaneously, all team members must still walk the same path for all teams. 

Thinking Strategically

When walking a path in Phase 2, you will be wearing body sensors from multiple teams, and is unlikely to be able to fool all of them.  Instead, you will need to choose your walking strategies carefully to fool the teams that are most likely to beat you, given the scores after Phase 1. The position estimates from each team will be shown live on a wall projection during Phase 1, as shown in the Figure to the right. This is an opportunity for each team to discern how the other systems work, and how they might fail. For example, radar and PIR sensors may not detect slow moving objects; accelerometers may be fooled by shaking or jumping, gyroscopes may be fooled by bending sideways, and so on. This knowledge can then be used during Phase 2 to score points for your own team but not for the other teams. No competitive strategies are off-limits, except that all players must walk monotonically forward along the predefined path. This means that running or jumping to fool body sensors or spraying baby powder to illuminate laser beams are all acceptable. Be sure to focus on avoiding the sensors of the team whose score is closest to yours.


On Game Day, each system must automatically (ie. no human intervention) report scores to a scoring server at least once per minute when it is in use. This can be done through the following XML-RPC function:

Reporting API

bool report(float X, float Y)

This function can be used to the current position estimate of a particular team.
The units of X and Y should be meters.

This function is provided by the competition server, for which the URL will be released on Game Day. In the meantime, the source code of the server is publicly available for testing. The code includes a XML-RPC server, a graphical visualization, and sample clients written in both Python and Java. Because this is a XML-RPC server, teams may call the function from any language that provides a XML-RPC library, including C, C++, Java, Perl, and Python, among others. See this simple "How To" to learn how to use a XML-RPC client in other languages. Teams will use HTTP authentication with the username and password provided.

To estimate the final position, teams should assume each path starts at (0,0) and can use a coordinate system with any orientation, as long as the units of the X and Y coordinates are meters. Thus, position estimates will be scored on their magnitude |X,Y| = sqrt(X^2+Y^2). In other words, position estimates are scored based on how closely they approximate the displacement of the final position form the original position. This score requires a team to differentiate between walking away from the origin and walking in a circle back to the origin, but does not penalize for the orientation of the team's coordinate system.

How to Enter

To enter, a team must send email to the competition co-chairs Kamin Whitehouse and Brano Kusy containing the team's name, institutional affiliation, the players' names, and a rough description of the hardware to be used. Entries must be received on or before April 5, 2009. The team should expect a response with the username and password to be used on the competition day. Email should be sent to:

In the case of any dispute, decisions about scoring and eligibility will be made by judges on Game Day and will be final. Contestants who would like to do something that might be the cause of dispute should ask for clarification of the rules and/or pre-approval of the hardware and techniques before the April 5 deadline. All rules of this contest are subject to clarification or change at any time.

* Wireless networks may be used and power cables of any length may be used.  One long data cable (e.g. a serial or USB cable) may be used to connect your sensor system to a computer.  However, no other cables that carry data may be more than 12 inches in length.

Travel Awards

The Extreme Sensing Competition organizers anticipate availability of Student Travel Grants to enable increased student participation in the competition. Each grant includes partial reimbursement for registration and travel costs.

  • Application Deadline: April 5, 2009


Student selection for the travel grants will be in accordance with the following guidelines:

  • Students who have already received another travel grant from an CPSWeek -related conference or workshop should not apply.
  • Travel grants will be awarded according to a financial need basis. Students who are well-funded and have ample means of covering their travel should not apply.

How to Apply

Interested student applicants should pay for the registration and attend CPSWeek first . If they are selected to receive the travel award, the travel award should cover partial expense of the students for registration and travel costs .

The student applicant must send an e-mail to The subject of the e-mail must be XSC TRAVEL GRANT APPLICATION (in uppercase letters). This e-mail must include the student's faculty advisor in the cc list. The e-mail must include the following information:




e-mail :

Advisor Name:

Research Topic:

Degree Sought:

Is your research funded by an active grant (yes/no) :
If Yes, is the source of funding NSF? (yes/no):

Underrepresented state/institution (yes/no):

Underrepresented/minority student (explain in 2-5 words):

The applications will be examined and ranked. The top applicants will receive award letters (by e-mail) informing them of the acceptance decision. They will then be instructed how to receive partial refunds, which will be issued to the person or institution that covered the initial expenses.

Women and members of underrepresented groups are particularly encouraged to apply. Applications from underrepresented groups will be given special consideration. Whereas most applicants are likely to be graduate students, undergraduates who are actively involved in systems research projects are also welcome to apply.

Student travel awards are supported by the National Science Foundation under Grant No. 0936902.

Other Notes

** A single-pixel sensor can be defined in many ways. As a general guideline, a "pixel" is a sensor input that is directly mapped to exactly one sensor output, with no logic in between. In some cases, a single AND or OR gate on binary sensors may be reasonable (in order to make the sensor bigger, for example). Single-pixel sensors that emulate multi-pixel sensors through movement of the sensor, the use of moving mirrors, etc, will not be acceptable. Similarly, using actuators that help the sensor emulate a multi-pixel sensor (such as a time synchronized array of infrared transmitters) will also not be acceptable. Borderline and questionable hardware should be discussed before the April 5 deadline, and all hardware is subject to approval by the judging panel on Game Day.