In large metropoles such as London or Delhi, the underground transportation system is the biggest energy consumer of the city, using as much electricity as 250,000 households. Over 30% of this energy is used by lighting, ventilation, climate control, security, information and communication, ticketing, and signalling systems in and around the underground stations. Although energy efficiency is internationally high on the political agenda, little is known about day-to-day energy consumption in these massive transportation hubs. Through empirical modelling of energy consumption and designing distributed control mechanisms for the metro station, the SEAM4US project was expected to reduce energy consumption by 5% – 10%. The Barcelona metro system has been used as a live test grounds.
The SEAM4US project proposed several steps towards reducing energy consumption. First, analysis would point out which systems are wasting energy under real-world conditions. Energy consumption by these systems has been empirically modelled during and off peak hours and under different daily and seasonal weather conditions.
In order to increase energy savings for such systems, users were encouraged to use the station in an energy-saving way. During off-peak hours, passengers could for instance be guided towards the main areas, so that in the auxiliary areas the lights could be dimmed, or elevators and ticket machines shut down. Of course, these areas had to be monitored for increased use, so passenger occupancy and flow models had to be acquired.
Another way to improve energy efficiency in the station’s systems, for example for ventilation, is exploiting external factors, such as outside temperature, wind or air flow from moving trains. These form renewable energy sources, which make overall climate and air quality operations cheap and low-power. In order to take full advantage of such factors, they must be predicted by empirical models, and a sensor-actuator network must be in place to open or close doors and windows.
Almende has designed and developed the passenger occupancy monitoring system via advanced computer vision algorithms performed over the videos taken by the existing CCTV infrastructure. The company also took care of the deployment of such system in the pilot station and integration with the rest of the SEAM4US architecture. Besides that, Almende contributed to the development of the intelligent control algorithms of SEAM4US and to the development and deployment of the overall SEAM4US architecture.
The project contributed to our expertise in working with computer vision algorithms to draw meaningful patterns from raw image data, as well as our expertise functioning as a system integrator. Finally, the project improved our knowledge on the development of data-driven control algorithms.