SP 3Concepts and pilot applications

Automated driving will come to prevail in the city when the technology can successfully demonstrate that it is able to master both junctions such as intersections and roundabouts and the challenges brought by traffic situations on urban thoroughfares. Furthermore, a solution must be found for the challenges involved in communication and collaboration with other road users. Most automated driving functions for applications in urban areas currently only exist as ideas. No fully developed concepts exist for scenarios such as driving through junctions and along thoroughfares, or for interacting with vulnerable road users.

Specifications and concepts

This subproject concentrates on developing the specifications and concepts for all these application scenarios in order to lay the corresponding foundations for the work being done in the other subprojects. Among other activities, this includes defining a catalogue of usage scenarios. The functional requirements,

research issues and criteria this results in, along with the specific scenarios derived from these, will be developed in the subproject right at the start of the project. Specifications that are as detailed as possible, covering all situations that need to be factored in and the required system performance, need to be defined for junctions – which include intersections and roundabouts.

To do this, all potential static topologies of a junction need to be systematically broken down in order to be able to represent any given junction as a composition of individual elements. All road users that need to be factored in (this includes, for example, vehicles used for public transport) must be defined in order to ensure that knowledge of all potential traffic situations occurring at the junction is available, as far as possible. Moreover, the preferred driving behaviour must be defined, and exact functional characteristics must be determined. The system specifications for automated driving through junctions are to form the basis for the derivation of specifications and requirements that apply, for example, to environmental sensing and to digital maps. The functional characteristics also require the human-vehicle interface to interact with the driver and other road users, for which there must be specifications as well. Characteristic scenarios for static bottlenecks are presented by narrowed lanes, structural measures to calm traffic which divide traffic lanes,

urban construction sites or delivery vehicles that park partially in lanes intended for traffic. The focus of the work is, however, directed at dynamic bottlenecks. Each static bottleneck becomes a dynamic bottleneck when, for example, aspects such as oncoming traffic, cyclists, car doors that open etc. are factored into the system configuration.

Automated vehicles will face the challenge of identifying and understanding the behaviour and intentions of other road users. This applies to a particularly high degree in situations involving vulnerable road users. Vulnerable road users often communicate in road traffic by using both conscious and unconscious poses or gestures. This could, for example, be the direction in which a pedestrian looks or the hand signal given by a cyclist when turning. To ensure an automated vehicle reacts appropriately when interacting with vulnerable road users, these characteristics of behaviour must be identified and be factored in to behavioural modelling and predictions.

‘Dynamic bottleneck’ pilot application

Along with the development of specifications and concepts, this subproject also concentrates on the implementation of functions in relation to situational awareness, the driving strategy and interaction, along with the integration and trialling of test vehicles.

Functional implementation of ‘situational awareness’ aims to generate a representation of the situation, which is used as a basis to plan and realise a suitable driving strategy for the ‘dynamic bottleneck’ pilot application. To do so, an environmental sensing module is being developed

that registers the urban environment and interprets it accordingly. The use of HD digital maps provides support with the environmental interpretation by making a priori knowledge, for example of the road infrastructure, available.

The objective of implementing the ‘driving strategy and interaction’ function is the functional implementation of all scenarios and requirements specified for automated driving in the form of pilot applications. In particular, the system should implement driving and interaction strategies that can deal with dynamic bottlenecks. This includes

delivery vehicles that, in part, park in lanes intended for traffic, but also cyclists who signalise their intent to turn a corner by making a gesture with their hand.

Ultimately, an initial system functionality (in relation to the ‘dynamic bottleneck’ pilot application) for automated driving along inner-city thoroughfares can be implemented on a test vehicle in an urban environment. This also includes a testing methodology with a testing specification, and the tools to be used for the tests.