ADVISOR is an automated visual surveillance system for metro stations which was developed as part of the project ADVISOR, involving 3 academic and 3 industrial project partners. The ADVISOR system aims at making public transport safer by automatically detecting at an early stage dangerous situations which may lead to accidents, violence or vandalism. In order to achieve this people are tracked across the station and their behaviours analysed. Additional measurements on crowd density and movement are also obtained. Warnings are generated and displayed to human operators for possible intervention. The article explores the main difficulties encountered during the design and implementation of ADVISOR and describes the ways in which they were solved. A prototype system has been built and extensively tested, proving the feasibility of automated visual surveillance systems. An analysis of test runs at a metro station in Barcelona and several individual experiments show that the system copes with many difficult image analysis problems. The analysis also points the way for future development and ways of deployment of the techniques used in the system.