June 2004

Grant Results

SUMMARY

From January 2002 to October 2003, researchers from the Massachusetts Institute of Technology (MIT) created a new automated technology for counting stair users.

They used this technology to test whether "just-in-time" health messages can motivate commuters to increase their routine physical activity by using stairs rather than escalators.

Key Results
Under the project, the researchers:

  • Developed a new computer vision algorithm that detects moving objects in public spaces and then calculates the percentages of people using either stairs or escalators.
  • Obtained approval from the Metropolitan Boston Transit Authority to test the new technology and conduct just-in-time messaging experiments in three Boston commuter rail stations.

Key Findings
The researchers reported the following findings to the Robert Wood Johnson Foundation (RWJF):

  • The new people-counting technology accurately calculated the percentage of stair users versus escalator users, adapted well to the difficult environment of the transit stations and was cost-effective.
  • When researchers used their new technology to project a motivational message ("Your heart needs exercise, here's your chance") in the transit stations, commuters increased their stair use by 4.3 percent, a finding consistent with prior studies.

Funding
RWJF provided $44,439 from February 2002 to July 2003 to support the pilot study.

 See Grant Detail & Contact Information
 Back to the Table of Contents


THE PROBLEM

According to a 1996 Surgeon General's report, only 22 percent of Americans are active enough to enjoy the health benefits of physical activity, a factor contributing to epidemic levels of obesity in the United States. Stair climbing can be an excellent form of physical activity.

In 1980, K.D. Brownell and others reported a study in the American Journal of Psychiatry that the placement of signs with health-related messages in public spaces led to a substantial increase in stair use by commuters. Researchers needed to study how best to use these "just-in-time" messages to increase physical activity, but the necessity of conducting time-intensive and expensive manual counts of stair users has hampered their investigation.

 Back to the Table of Contents


RWJF STRATEGY

This project relates to RWJF's objective of increasing physical activity and reducing the epidemic of obesity in the United States. It complements RWJF's national program, Active Living by Design, a $16.5 million national program designed to establish and evaluate innovative environmental approaches to motivating people to accumulate at least thirty minutes of physical activity each day.

 Back to the Table of Contents


THE PROJECT

The objectives of the pilot study were: (1) to create an affordable, automated people-counting technology to replace labor intensive manual observation of stair users; and (2) to use this technology to demonstrate how interactive, digital messages can motivate and sustain stair use in public spaces. The Project Director was Stephen Intille, Ph.D., Director of MIT's House — a multidisciplinary project in the School of Architecture and Planning, to build technologies and environments that promote more active lifestyles.

To replace manual counting, the researchers modified existing computer software, writing new computer vision algorithms that detect moving objects in public spaces and then calculate the percentages of people using either stairs or escalators.

The researchers obtained approval from the Metropolitan Boston Transit Authority to test the new technology and conduct just-in-time messaging experiments in three Boston commuter rail stations. By summer 2003, the researchers completed installation of their equipment, consisting of portable projectors that beamed messages onto blank signs and a laptop computer with a small attached camera that continuously ran the new algorithms and collected data on stair users.

Due to extended testing of the algorithms and delays in installing their equipment, the researchers completed only one "just-in-time" messaging study before the grant closed and they were unable to test the impact of different interactive messages as originally proposed.

 Back to the Table of Contents


FINDINGS

The researchers described their findings in a report to RWJF:

  • The new people-counting technology accurately calculated the percentage of stair users versus escalator users. The researchers manually counted 24,186 people exiting transit stations using the escalator and 10,253 people exiting using the stairs and compared their results to the algorithm's performance on the same data set. The difference between calculated stair use percentage and actual percentage ranged from -3.8 percent to 2.8 percent. The researchers concluded that the new measurement tool was sufficiently accurate for use in future stair use studies.
  • The new technology adapted well to the difficult environment of the transit stations. Unlike prior people-counting systems, the new technology was able to handle the large crowds moving through the transit station and functioned properly in dim and changing lighting.
  • The new technology proved cost-effective. The computer vision algorithm can run continuously, collecting substantially more observations than previous methods. Over a sixteen-day period, the new system collected 46,442 observations of commuters before and after a motivational message appeared in the transit stations.
  • When researchers used their new technology to project a motivational message ("Your heart needs exercise, here's your chance") in the transit stations, commuters increased their stair use by 4.3 percent, a finding consistent with prior studies. During the nine-day intervention period, 43.7 percent of commuters chose to use the stairs compared to 39.3 percent during the seven-day baseline period when the sign was not visible.

Communications

The researchers submitted a report on the development and evaluation of the new technology, Computer Vision Based People Tracking for Motivating Behavior in Public Spaces, to RWJF and MIT. They will continue to communicate the results of ongoing experiments to RWJF and will publish their findings through the popular media, academic journals, and an MIT Web site. See the Bibliography for detailed information.

 Back to the Table of Contents


SIGNIFICANCE TO THE FIELD

The preliminary research using the new technology confirmed findings of Brownell and others. However, experiments conducted since the RWJF grant closed suggest that, after removal of the motivational signs, commuters return to their previous stair use levels sooner than prior studies had indicated.

 Back to the Table of Contents


LESSONS LEARNED

In a report to RWJF, the Project Director cited the following lessons learned from this pilot project:

  1. When studying behavior in public spaces, researchers should beware of publicizing their work too early. The MIT name on the transit station signs attracted the attention of newspaper reporters, who wanted to run an article on the stair use study. The researchers declined, fearing that premature publicity might bias their results by increasing the public's awareness of the signs and influencing their behavior. (Project Director/Intille)
  2. Logistical complexities can substantially increase the time and cost of conducting natural experiments in public places. Although the Metropolitan Boston Transit Authority was highly cooperative, frequent personnel changes and a lengthy approval process for installing the equipment seriously delayed project implementation. (Project Director/Intille)

 Back to the Table of Contents


AFTER THE GRANT

The MIT researchers continue to collect data on stair use at the three transit station locations. They are using the counting technology to study whether animated signs that change in response to the environment have a more lasting impact on stair use than static signs.

 Back to the Table of Contents


GRANT DETAILS & CONTACT INFORMATION

Project

Improving Methods for Measuring Stair Use in Public Spaces

Grantee

Massachusetts Institute of Technology (Cambridge,  MA)

  • Amount: $ 44,439
    Dates: February 2002 to July 2003
    ID#:  044256

Contact

Stephen S. Intille
(617) 452-2346
intille@mit.edu

 Back to the Table of Contents


BIBLIOGRAPHY

(Current as of date of this report; as provided by grantee organization; not verified by RWJF; items not available from RWJF.)

Reports

Hyman JA. Computer Vision Based People Tracking for Motivating Behavior in Public Spaces. Cambridge, Mass.: Massachusetts Institute of Technology, 2003. Submitted to the Department of Electrical Engineering and Computer Science in partial fulfillment of the requirements for the degree of Master of Engineering in Electrical Engineering and Computer Science.

Audio-Visuals and Computer Software

Software for Estimating Stair Versus Escalator Use, a CD-ROM containing data collected from the three stair use study sites, code documentation, and "getting started" information on using the counting technology. Cambridge, Mass.: Massachusetts Institute of Technology, 2003.

Data Tapes

Measuring and Motivating Stair Use in Public Spaces, a CD-ROM including data collected using the counting technology developed under this project, consisting of stair counts from the three Boston transit station installation sites. Cambridge, Mass.: Massachusetts Institute of Technology, 2003–2004. (CD-ROM to be distributed to the Inter-university Consortium for Political and Social Research and updated as more data are collected.)

 Back to the Table of Contents


Report prepared by: Jayme Hannay
Reviewed by: Robert Crum
Reviewed by: Molly McKaughan
Program Officer: Terry L. Bazzare