University Hospital of Zealand Autonomous Bus

Movia, the Public Transport Authority (PTA) of Greater Copenhagen has planned three pilots to test autonomous busses on hospital sites. Phase 1 was an indoor, 3 month-long test with a bi-directional bus. The vehicle operated on-demand, at low speed (3.5km/h) and was wheelchair accessible. The bus had four coordinated based stops, circulated on a closed course and in a small space with a lot of pedestrians and indoor hospital vehicles. The test site was at the University Hospital of Zealand, Køge. This pilot was not covered by the Danish legislation for autonomous vehicle testing.

Movia, as a PTA, has policy priorities which seek to push the development of autonomous technology in public transport. They believe it is important to test autonomous buses interlinked with public transport in order to understand and fine-tune the technology in relation to topics such as user behavior, user-friendliness (e.g. disability and satisfaction), user acceptance and ordinary operation in the context of public transport.

The hospital autonomous bus projects have several objectives:

• Conduct autonomous passenger transport on three selected routes, with an increasing complexity;
• Collect knowledge and experience on the application process to obtain the authority approval, safety assessment, mobilisation, operation, evaluation and other relevant conditions for the further development and dissemination of the technology;
• Test the technology in different contexts and use cases (the three project phases).
• Extract and disseminate lessons learned for the benefit of municipalities and regions, relevant authorities and other public institutions, operators and suppliers, as well as public transport passengers.

The budget of phase 1 was estimated at €90,000. The budget of the three tests altogether is estimated at €605,000. The pilots are financed by the Region of Zealand (45%), Movia (45%) and the Capital Region of Copenhagen (10%). The Metro Company is also contributing with man-power.

The day-to-day operations of the autonomous shuttle for phase 1 of the project gave Movia valuable knowledge and experience. Challenges encountered were especially instructive. These include: the indoor setting which gave circumstantial challenges as the GPS signals was unreliable and put extra pressure on the image recognition to align the bus with the ramps (ramps were set in place to allow easy access to the bus for wheelchairs, walkers, disabled and not-so-sure-footed passengers at the hospital). The software systems therefore needed manual re-setting (re-calibration) in order to allow perfect alignment of the ramps. In theory, SAE level 4 should not have this issue. There were also difficulties with the communication between the vehicle and the on-demand system - manual redirecting of the bus to the desired stop was necessary. A much more effective communication tool is needed for passengers to understand – and use – the on-demand call of the bus stops. Although the bus operated in on-demand mode in the off-peak hours, most passengers just waited for the bus to come, rather than actively summoning it.

By the end of the trial, 5,800 passengers had been transported with an average of 75 people per day. The shuttle travelled a total of 842km (on average, 13km per day) at a typical operational speed of 3.6km/h. The customer satisfaction was very high - 98% users claimed to be satisfied with the service. No accident or incidents involving passengers occurred during the three-month trial. However, there were three disruptions in the performance of the bus: there was a breakdown of a shock absorber on the bus (this was fixed within one working day). The battery of the vehicle ran flat but this was due to a manual mistake (forgot to connect to the power socket). And finally there was a minor collision with a passenger ramp (at 1 km/h). This led to the need of a software update and a readjustment of the ramp.

Two studies were carried out to measure customer satisfaction; “Interaction with the build-up area” by the University of Aalborg and “Passenger perceptions” by The University of Roskilde.