In most developed countries, people with physical disabilities are entitled to special transportation services which are often dispatched on a pre-booked basis as a means to help supplement the existing fixed-route bus and rail networks. These types of services are delivered through a dedicated fleet of vehicles that operate in accordance with a complex pre-planning procedure. This is often made more problematic due to the nature of the trips; operating at sporadic times and serving multiple locations which in turn makes itineraries rather difficult to match.
The Empresa Municipal de Transportes (EMT) in Valencia is one of the largest public transport authorities in Spain. Earlier this year, EMT contacted Shotl as they wanted to find ways in which to improve some operational procedures relating to the services they were delivering. Their intention was to enhance the efficiency of these routes while providing a much more convenient service for users with physical disabilities.
Despite the fact that Valencia’s entire fleet is already wheelchair accessible, EMT has allocated a small fleet of buses to provide special transportation services to people whose disability might keep them from accessing regular buses. This fleet includes a total of four 12 meter long buses that are able to transport up to 6 passengers at once. Currently, users need to make a reservation for this service in advance, whether for recurrent trips throughout the year, such as daily commutes to school or to work or for occasional rides, in which case it is necessary to call 48 hours in advance.
So far, requests are answered and organized entirely by a full-time dispatcher whose job it is to combine all occasional rides with pre-scheduled trips. As itineraries have to be merged in accordance with requests, routes must be created daily. Unfortunately, when a ride gets canceled or a passenger does not show up, itineraries remain as they were previously designed, causing a loss in efficiency.
As per EMT's request, Shotl’s mission is to provide an on-demand bus dispatching system that's able to match all pre-scheduled rides up to 4 hours in advance and adapts to changes in on-demand fluctuations, cancellations, no-shows, and altered road conditions. All in real time.
The soft-launch started at the end of October 2018, this has allowed time for the algorithm to automatically design daily routes; considering trip types (recurrent or occasional), as well as longer pick-up and drop-off times in accordance with the requirements of each specific user. Gradually, the routes generated by Shotl have been replacing the ones created manually, and the dispatcher who was previously assigned for that purpose is now supervising the process. Service users have the option of being picked up and dropped off at either existing bus stops throughout the city or at their front door in cases where users might require special assistance.
The hard launch is set for January 2019, when all incoming phone calls will be differed to the general EMT call center, instead of a dedicated one. All agents will have access to a Shotl web dashboard where bookings can be processed easily. Also, for the first time, passengers will have access to the Shotl Passenger App which allows for bookings, changes, and cancellations of recurrent rides to be made via smartphone, with no need for a phone call.
Users will also be able to request a ride on the day of travel, instead of being restricted by the current 48 hour limit on pre-bookings. Minutes before being picked up, users will also be able to track the exact position of their assigned vehicle in real-time. If a ride gets canceled at the last minute, the Shotl algorithm will be able to fill the free slot with another trip.
The managing team at EMT Valencia will be able to track the position and the route of every vehicle and will receive weekly reports, including all KPIs previously defined by the transport operator, such as average speed, number of passengers per kilometre, average traveling time, average waiting time, etc.
The success of this collaboration will be measured both on an operational level and as per users' level of satisfaction. Shotl’s routing algorithms will help to save on mileage and resources whilst most importantly helping to make a positive impact on the lives of people with physical disabilities, who will enjoy a much more flexible service, that's more accessible, equitable and fair for all.
The Shotl team would like to thank EMT Valencia for their trust and willingness to work together on such a noble endeavour.
Stay tuned as we look forward to publishing the results of this project next July.
Machine Learning (ML) is a field of Artificial Intelligence (AI) which focuses on the development of techniques that allow computers to learn how to perform a specific task.