Public journey planning uses centralised computing power (servers) to find routes for users, which is costly and increases with the number of users. It also means a users journey is shared with the service provider so full data privacy is impossible.
We set a goal of designing a commodity journey planning service with baked in data privacy and very low costs to run. Our design process focused on using a decentralised network model putting a users mobile device at the core of the system design allowing us to lock all personal data locally.
Public transport journey planning is a well documented computing problem with many things to consider including service times, service speed, interconnection possibilities between services and different modes of transport. Consequently providing the most logical journey to match users expectations across metropolitan, suburban and rural transportation options is highly complex.
To decentralise this planning problem we needed to reduce it to something a user’s mobile device could solve as opposed to cloud computing; it was limited by the computing power, memory and storage of the device. CPU power directly affected speed, memory how many potential routes we could look at concurrently and storage affected how much data could be held locally.
Using novel software engineering we created a client-side software library, which can be embedded for example within a mobile app. The algorithm reduces complexity, data and the search for routes across an unlimited geographical area, with fast results. Static industry service data (e.g timetables) is managed in the cloud, collated and preprocessed to an optimal algorithmic structure, then provided to the app when data changes.
By employing a deep knowledge of software engineering, architecture and digital product design we met our goals of providing full data privacy in a very low cost commodity public transport planning system. Our approach provides a technological step change over that employed by existing providers of national and international door to door planning services e.g. Google, Apple, and Moovit. It also opens the way for new business models and user-centred design features that are otherwise not possible. We would like to thank Ayoupa for testing and using our technology in the iOS App: Commuter (downloadable from the App Store), and Innovate UK for backing the idea and providing a degree of funding support.