Problem
It is difficult to ascertain accurately and cost-effectively how many people are on a bus at any one time. While the latest contactless ticketing records passengers getting on, it doesn’t record passengers getting off (see our Smart Ticketing case study on Tap On, Tap Off innovation). Plymouth Citybus needed more detailed information on exact passenger patronage to be able to optimise route management and improve the customer experience. This data could also provide the basis for significant cost saving potential.
Solution
Partnering with a camera and artificial intelligence (AI) technology provider, a solution was developed to gather occupancy data. Cameras were deployed at ground level on the entry and exit points of buses on a selected route. The camera and the supporting AI were programmed to recognise and differentiate passengers by their footwear, eliminating any GDPR issues, and identify when they got on and off the bus. This information was relayed to a back-office system that was able to provide detailed route analysis of when and where passengers got on and off.
Results
Following an initial trial period, Plymouth Citybus has identified a number of areas where they can improve their operations. First is bus choice; the statistics have enabled the operator to decide on whether a single or double decker bus would be most appropriate for a particular route, which can reduce operational costs. Second is route optimisation; bus operators can adjust their routes and timetables based on actual use, which improves the customer experience. And, third, is bus procurement; this would have traditionally been influenced by high level modelling of expected bus use but now the decision on whether to buy a single or double decker bus is based on hard evidence. This also affects future investment strategy, which will enable Plymouth Citybus to significantly reduce its costs over the life of the service.
If you’d like to know more about AI foot recognition, please contact us.