Overview
Public Transportation organisations play a significant role in the everyday quality of life and provide their services to an ever increasing number of people. There is need for the improvement of the productivity and quality of these services. However, strictly improving a bus route productivity can have detrimental effects on the quality of service. Passengers need more buses with enough space to have a comfortable and safe travel to their destination. On the other hand, Public Transportation Companies have to discover optimised bus schedules, so that during operation no bus will be relatively empty or exceed a desired minimum number of passengers, leading to unprofitable lines and financial losses.
The main objective of the project was the design and implementation of a novel information system, BusGrid, for the administrators of Public Transport companies with a two-fold goal of improving both quality of service and productivity of bus routes.
The system produces solutions that tackle this trade-off by consistently optimizing vehicle schedules to achieve a relatively stable average number of on board passengers. Specifically, it receives real time data from installed sensors (AVL & APC) on the vehicles of Public Transportation Companies in order to:
- Calculate quality of service key performance indexes (KPIs) and improve bus line’s productivity
- Analyse and produce useful information which will assist the prediction of critical factors (such as bus demand in each bus stop)
- Actively support the decision process of bus routes improvement and optimal response to passenger demand
BusGrid is implemented as a series of interconnected modules. The data collection module gathers a series of data from the AVL and APC systems and other installed sensors such as the engine ignition on/off, door sensors etc. Using the bus’s wireless communication system (e.g. GSM), a unified data-frame containing the sensor data is sent to the database server in real-time. Next, a data layer is created by preprocessing the raw data of the database. This stage includes data cleaning, regularisation and aggregation of data (e.g. creating a per-day view of the data). The data-layer created, forms the data-sets used by the modules that calculate the KPIs of the bus routes and by the Machine Learning modules. The KPIs are calculated directly from the raw data.
The Machine Learning modules are used for the prediction of critical KPIs such as passenger demand and for the adaptation of the system in various cases. The first Machine Learning module (Prediction of Demand) estimates the demand for each specific bus route at each bus stop. It uses examples of temporal and geospatial data, combined with the historical data from the APC sensors, for the supervised learning of a regression model which predicts people demand for a specific bus stop, bus and period of time. The average waiting time module produces a KPI based on the actual average waiting time weighted by the passenger demand (number of people waiting) so that waiting time is expressed in man-hours.
Finally, the RL-BUS module uses Reinforcement Learning methods to dynamically create adaptive bus schedules optimised for bus productivity and quality of service. It is a semi-supervised model, learning a scheduling policy for dispatching buses on the correct time to achieve optimal bus productivity with respect to the average bus fullness and waiting time (quality of service indicators). The KPIs calculation modules and the Machine Learning modules are developed as software libraries, separating the interface tier of the system (GUI) from its business logic.
Funding
Results
The BusGrid project has produced four deliverables (reports) on the design and implementation of the software and a CD-ROM with the software itself.
The resulting system can produce:
- Accurate predictions for passenger demand
- Adaptive schedules based on the available fleet vehicles and the passenger demand models
- A series of informative KPIs along with illustrative plots and diagrams for the current productivity and demand status of the bus-fleet.
Innovation aspects
BusGrid delivers high prediction accuracy and optimized bus schedules based on a novel computational representation of the problem and the chosen Machine Learning methods. More details can be found in the following publication: https://dl.acm.org/citation.cfm?id=2801984
Policy implications
Information Systems such as BusGrid promote the design and use of new informed and demand-driven public transport policies. Adopting such policies can significantly improve the quality of service for the passengers, while also contributing to an efficient and cost-sensitive use of resources, such as the vehicles of any public transport fleet.
Readiness
The system can be used in any Public Transportation Company after its necessary custom adaptation to the different sensors and data used.