Final Evaluation Report

Executive Summary

ADVISOR is a project that intends to improve management of public transport networks through the improved use of CCTV cameras. The project undertakes to integrate several developments in picture content detection and analysis with technologies for annotating video images, and presenting them for inspection by CCTV operators when trigger events occur.

In other words, the goal of ADVISOR is to assist human operators by automatic selection, recording and annotation of "interesting" images such as "abnormal" crowd and individual's behaviours. Metro stations have been used as experimentation sites for test beds (providing video recordings) and final demonstrator (on site installation).

Three main evaluation milestones have been programmed during the project's life : after the completion of Test Bed 1, Test Bed 2 and Final Demonstration prototypes respectively. They were conducted according to the Evaluation Plan previously set-up and described in Deliverable R8.1.

The present document reports the final evaluation performed by the end-users associated with the project, i.e. STIB (metro of Brussels) and TMB (metro of Barcelona) during the one week evaluation session that was organised in March 2003 at the central control room of TMB for operators and representatives from Brussels and Barcelona.

The Final demonstrator took inputs from both live cameras from a station of the Barcelona metro network and pre-recorded video collected from multiple cameras of sites provided by TMB (Barcelona) and STIB (Brussels).

The ADVISOR final prototype had the following four main functions :

• Capture and Video Streaming
• Machine Vision Processing (Motion Detector + Crowd Monitor + Behaviour Recognition)
• Archive, Search and Retrieval
• Human Computer Interface

It is through the Human Computer Interface (HCI) that the ADVISOR system has been made available to End-users. This HCI has allowed them to activate some functions and also to be alerted by the system as new interesting events or incidents are detected. At this final level of development of the project, most of the functionality had been implemented, which allowed complete demonstration and credible evaluation.

The following events/situations have been used for the final evaluation :

• Interesting crowd movements
  • overcrowding
  • blocking entry/exit
• Individual delinquency
  • violence
  • people without ticket (jumping over the barriers)
  • vandalism

The evaluators have been in a position to verify how the ADVISOR system detects such situations and their evaluation has been both quantitative (when relevant) and qualitative.

Three different aspects of the system have been evaluated by the End-users, i.e. the "user acceptance", the "impact" and the "socio-economic consequences", each of the aspects being managed as follows :

• User acceptance evaluation
  • Usefulness of the system
  • Operating the system
  • Using the system
• Impact evaluation
  • Impact on safety
  • Impact on security
  • Impact on traffic efficiency
  • Impact on personnel motivation
  • Impact on passengers behaviour
• Socio-economic evaluation in which we provide a calculation of implementation costs and we propose to compare these costs with the potential benefits
  • Benefit gained from the increased number of passengers
  • Other social benefits

The final evaluation reported in the present document describes how the ADVISOR system has been very well appreciated by the relevant potential users, i.e. the metro companies.

They have acknowledged the efficiency of the ADVISOR system and confirmed that it could help them to increase the efficiency of their work. They recognised that many of its functions could not be provided by any other existing system, which demonstrates its innovative and advanced nature. This definitely confirms the success of the ADVISOR project.