With the support of the Ministry of Transport, Public Works and Water Management, in 2005/2006 NS and ProRail carried out the Innovative Experiments Social Safety (INES) project. During INES they investigated how innovative techniques contribute to the safety and perception of safety of travellers and NS employees. TNO fulfilled a variety of roles during the project, from market research and benchmarking to supervising the field test.
An initial question that NS and ProRail posed was whether intelligent systems are available that detect incidents such as items being left behind, people climbing over gates, individuals on the tracks and pickpockets at work. Any such intelligent system would also need to be capable of verifying a persons presence by a Service and Alarm column and of recognising crowds or queues at the foot of escalators. In addition, the partners went in search of systems that could support the retrospective shifting of video data. TNO approached various companies in the Netherlands and abroad with the question whether they had intelligent camera systems that met the requirements of NS and ProRail. Based on their answers, it was decided to invite ten companies for a second round of questioning. At this time, they demonstrated their capabilities using visual material staged by NS/ProRail and TNO. TNOs thorough knowledge in the field of video analysis helped the parties evaluate the different systems.
Three companies were eventually selected to implement their systems in the experimental environment created by NS/ProRail. These systems were evaluated over a couple of months. TNO was called in during this phase to design the technical test plans. This meant that each company knew which data their system needed to supply. One element of the test plans was a variety of test scenarios.These were paid special attention during the trial period. In principle, very many tests could be performed by NS staff. In these tests, TNO´s role was mainly supervisory. Others of the tests were carried out by TNO. In both cases, it was important that TNO had developed the correct test scenarios by drawing on a good understanding of the operation of the new technology. This enabled the value of the systems to be determined in practice.
In its first phase, the project for NS and ProRail afforded an understanding of what is currently state of the art in intelligent camera systems. For example, it is evidently not yet possible to automatically detect pickpockets at work. Neither was any supplier able to demonstrate the consecutive use of several cameras to track an individual´s movements. By contrast, there was plenty of choice when it came to systems that detect incidents such as items being left behind and individuals climbing over fences. During the field test it became apparent that very many innovative technologies could not be included in an NS and ProRail security concept without adaptation. For example, the dynamic and busy station environment poses problems for intelligent camera systems. As the technology involved is relatively new, a complex environment like that used in the INES tests reveals many problems for the first time. TNO helps identify and evaluate these problems.