Introduction



Abstract

The introduction of Automatic Intersection Management (AIM) promises higher traffic flows and air quality by increasing the effective capacity of the intersection. Unfortunately, automatic control is only possible in a fully autonomous environment without legacy vehicles and vulnerable road users like cyclists and pedestrians. In this paper we introduce Hybrid Intersection Management: a concept which makes it possible to gain the benefits from autonomous vehicles in the situation with mixed traffic. A new message protocol, the individual Space & Time message (iSPaT), is introduced to effectively send instructions to autonomous vehicles. The iSPaT message is also used to guide autonomous vehicles safely over the intersection in a ‘woven’ manner. The development of Hybrid Intersection Management can be started today, and shifts towards automatic control when the amount of autonomous vehicles increases.

Keywords

Automatic Intersection Management (AIM), Cooperative Automatic Vehicles (CAVs), mixed traffic.

Introduction

In the far future of automated and cooperative driving the question rises whether traffic lights at intersections are still necessary. Well-known visions sketch the situation of traffic crossing the intersection in a ‘woven’ manner supported by Autonomous Intersection Management (AIM). From the nowadays green periods, a transition is made towards very adaptive and flexible individual, automated vehicle-bound, occupations of space and time.

Before we reach this far future we will have to find solutions for the reality of a transition phase, in which traffic is partly 'legacy' and/or disconnected and partly 'Cooperative Automated Vehicles' (CAV’s). Legacy traffic is a mix of motorized and vulnerable traffic, like pedestrians and cyclists. This adds to the complexity of mixed traffic in and urban environment and leads to the term of a ‘hybrid intersection’: an intersection where CAV’s, regular motorized vehicles and vulnerable road users compete for capacity.

Other trends within mobility are concepts like MaaS and the sharing economy. These trends make it difficult to assess what the future mobility system will look like. (Long term) economic growth and MaaS are expected to greatly increase the number of travelled kilometres adding to the need for a higher capacity on intersections.

This paper focuses on an urban environment and urban intersections. The paper analyses a practical approach to enable automated and legacy vehicles on a hybrid intersection in an urban context within a transition phase of 10 to 20 years, from our current legacy traffic system towards a system with CAV's as a given. For this transition phase studies have been performed to assess how this mix can be handled.

Looking at simulations of these future intersections three key concerns arise.


  • First, there will be a transition period from the current system towards a traffic system in which all traffic is automated. In this transition period current vehicles will be mixed with ‘cooperative automated Vehicles’ (CAV’s). CAV’s are vehicles that drive autonomous or at least highly automated and are connected with their environment, so that cooperation is possible. Especially in urban setting this gives a challenging situation because space is scarce, the traffic situation is complex and the availability of reliable data of CAV’s in this transition phase is a challenge. Also in the future scenario that all vehicles are CAV’s, it is likely that data is splintered among stakeholders.


  • The second concern is the extra complexity of the interrelated space and time reservations. On an intersection with numerous vehicles of multiple brands and types the possibility of a successful continuous group negotiation to organize these interrelated crossings can be expected to be close to zero. Next to that the group of vehicles lacks the (anticipatory) insight on an urban scale to optimize traffic flows starting from a larger distance and the group of vehicles cannot be responsible for assigning priorities based on policy.


  • The third concern is for vulnerable road users (VRU’s) like pedestrians and cyclists, they will either have no chance of crossing unless some kind of supervisor claims space and time for them; or they cross anyway expecting the AV’s to stop reducing the efficiency or blocking the traffic flow. It is unlikely that this is acceptable from the local policy viewpoint. A kind of supervisor will require input to detect these road users and requires a way of communicating the space and time reservation towards these users.


Above concerns come together within traffic management where a public authority is responsible for safe, fast and reliable journeys. This responsibility is translated in policies and priorities for types of traffic.

The shift towards a system in which traffic management and control are autonomous as well as the vehicles, is complex from a functional and technical perspective. Within this paper we perform an analysis on a technical level and come to a proposal to enable AIM and facilitate the transition towards the fully automated future.

COEN BRESSER

Senior consultant Smart Mobility

BAS VAN DER BIJL

Innovation manager