The project delivers a modular software architecture that enables on‑demand public transport (ODV) in rural and peri‑urban areas. Building on the open‑source OpenTripPlanner (OTP) framework, the architecture incorporates four key extensions identified during the requirements workshop. First, the system adopts GTFS‑Flex v2 to generate dynamic timetables in near real‑time, allowing flexible services that respond to individual passenger requests. Second, a routing algorithm developed by the Braunschweig University of Technology (BTU) is integrated into OTP’s routing API, replacing the default Dijkstra or A* routines with a more efficient, multi‑modal planner that can handle short cross‑connections and evening off‑peak trips. Third, an interface is defined for connecting external booking systems, enabling passengers to reserve rides through existing ticketing platforms. Fourth, a dedicated API links the system to the bbnavi navigation app, allowing seamless hand‑off between the ODV service and the user’s mobile device.

Data ingestion is comprehensive: OTP pulls GTFS feeds from Berlin and Brandenburg transit operators, OpenStreetMap for street networks, and GBFS feeds for bike‑share locations. The GTFS‑Flex v2 feed adds two new extensions—FlexibleTrip and BookingRules—that encode service flexibility and booking constraints. By generating a transit feed that conforms to this specification, the architecture can produce a transit graph that includes both fixed‑route and on‑demand services. The graph is then queried through OTP’s GraphQL API, providing real‑time route calculations that consider distance, travel time, transfers, and walking distances. The integration of the BTU routing algorithm ensures that the planner can propose efficient pickup points, minimize detours, and maximize vehicle occupancy, especially for short cross‑connections and evening trips where conventional fixed routes are unavailable.

Three application scenarios were evaluated. In the commuting case, the system optimises transfers between modes and guarantees a fixed arrival time, suitable for regular weekday travel. For short cross‑connections, the focus shifts to flexible booking and efficient pickup points, handling spontaneous requests that cannot be served by existing lines. The evening off‑peak scenario addresses the lack of public transport after 21:30 h; the ODV service bridges the gap between a central restaurant and peripheral districts, reducing car travel from an average of 12 minutes to a 29‑minute bus ride or a 1 hour 24 minute transfer. In a town the size of Spremberg, the study estimates that one to two vehicles would suffice for these use cases, with scalability to larger fleets as demand grows.

Collaboration among partners underpins the technical achievements. The Brandenburg Technical University (THB) led the requirement definition and quality assurance, ensuring that the architecture meets non‑functional criteria such as reliability, extensibility, security, and data privacy. BTU supplied the custom routing algorithm and contributed to the integration strategy. neuland21 performed the data collection and assessment of rural transport conditions, providing the empirical foundation for the system design. The project was carried out over a multi‑year period funded by German federal and European Union programmes that support digital mobility solutions. Together, the partners established a reusable, open‑source framework that can be adapted to other regions seeking to enhance mobility through on‑demand public transport.