Chippendale Automated Vehicle Trial

The Chippendale Automated Vehicle Trial has been completed. In addition to investigating the introduction of Connected and Automated Vehicles (CAVs) into the network, it also looked to understand the interactions between CAVs, vulnerable road users (VRU) and other road occupants in various scenarios and contexts. This project will also facilitate the future study of how CAVs interact with other traffic and infrastructure such as traffic lights, other connected vehicles, and smart infrastructure.

Working with iMOVE on the project were Transport for NSW and the University of Sydney’s Australian Centre for Robotics (ACFR). Work on this project was completed in mid-2023.

Project aims

The main aims of the project were:

• Develop recommendations for safe CAVs to operations in complex and non-complex environments, and as part of an integrated transport network
• Assess the impacts of CAVs when operating in different environments with special attention placed on high density pedestrian areas
• Establish a number of real-world testbed sites for CAVs in varying complexity environments for further research in evaluating emerging CAV and Cooperative Intelligent Transportation Systems (C-ITS) technology
• Improve industry understanding of the physical and digital infrastructure needed to support the safe introduction of CAVs in NSW

The test vehicle

The vehicle used for this project was a Level 4 capable autonomous vehicle, purpose-built in 2017 and it has been used for numerous pieces of autonomy research.

Sensors aboard the vehicle were:

• 1 x multi-beam scanning LiDAR (Light Detection and Ranging) sensor with a 30◦ vertical Field-of-View (FoV) and a 360°horizontal FoV for perception, low-level (local) localisation and navigation.
• 6 x Gigabit Multimedia Serial Link cameras with 1080p resolution and a 100°horizontal FoV each – positioned such that they may be stitched together to provide a 360°horizontal FoV.
• 1 x Cohda Wireless MK5 On-Board Unit transceiver for communicating in Vehicle-to-Everything (V2X) networks.
• 1 x Global Navigation Satellite System for high-level (global) localisation.
• 1 x 6 Degrees of Freedom Inertial Measurement Unit for odometry.
• 4 x wheel encoders for measuring the rotational speed of the wheels.
• 1 x wireless Hetronic e-stop transmitter and receiver for entering/exiting autonomous mode

The vehicle also included a system of external lights to convey the vehicles intent and awareness of other road users, particularly pedestrians.

Transport for NSW provided regulatory support to approve the vehicle to operate on approved NSW roads.

Project phases

In order to accomplish its aims, the project was divided into three phases, each in a different test environment, and in increasing levels of complexity.

• Phase 1: Low density urban / pedestrian environment – Focusing on pedestrian detection and interactions with other vehicles in a controlled manner.

The maximum drive duration for this phase was 10 minutes, and the maximum speed was 12 km/h. The road used were public, a 250-metre loop within the grounds of the University of Sydney.

• Phase 2: Mixture of low-speed urban and pedestrian environments – Operating in low-speed urban environment using Intelligent Roadside Infrastructure for safety and traffic coordination, building cooperative Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communications.

The maximum drive duration for this phase was 30 minutes, and the maximum speed was 10 km/h in shared space zones, and 20 km/h on public roads. The road used were public, a 900-metre route on the university campus.

• Phase 3: Higher-speed urban and pedestrian environments – Operating in urban environments to develop and trial Autonomous Transit on Demand (ATOD) and safe transitions through several driving environments.

The maximum drive duration for this phase was 30 minutes, and the maximum speed was 10 km/h in shared space zones, and 25 km/h on public roads. The road used were public, a 700-metre route on the university campus and out onto public roads that included Meagher, Balfour, Abercrombie, Myrtle, and Smithers Streets, Chippendale.  This route included signalised and unsignalised intersections, a roundabout, and traffic lights.

Off-road testing was conducted at the Future Mobility Testing and Research Centre in Cudal prior to operating on public roads.

Trial outcomes

Development of CAV guidelines and recommendations: Based upon the three phases of trial at the University of Sydney campus and surrounding Chippendale for the safe operation of connected and automated vehicles in complex environments and as part of an integrated transport network were provided to Transport for NSW. Key guidelines and recommendations can be found in the final trial report.

Assessment of CAV operation in high-density pedestrian areas: The ability to demonstrate and assess the impacts and benefits of  in various contexts, particularly in densely populated pedestrian areas, was hindered by the decrease in pedestrian numbers caused by the COVID pandemic. To address this, researchers utilised existing data and inputted it into the software to simulate the decisions made by the vehicle systems in high pedestrian scenarios. This enabled researchers to visualise and understand vehicle decisions more effectively in real-world scenarios.

Establishment of real-world CAV testbeds: A key trial requirement was the establishment of a number of real-world  testbeds. In addition to the testing environments in and around Chippendale, as part of this project another new testbed was established at Transport for NSW’s Future Mobility Testing and Research Centre (FMTRC) in Cudal, a town near Orange in Central West NSW. This additional testbed allowed leads to understand CAV operation in regional environments.

Increase industry understanding of CAVs: The development of automated vehicle technology and software by the ACFR team for this project eliminated the typical constraints and compromises associated with off-the-shelf commercial solutions. The findings from this project enables industry to improve their understanding of the inner workings of CAV technology.

Next steps

Transport for NSW is continuing to work with the University of Sydney to further understand emerging connected and automated vehicle technologies and their ability to deliver a safer, more robust transport ecosystem.

An extended CAV trial using C-ITS modified infrastructure, focusing on vehicle to infrastructure (V2I) and vehicle to vehicle (V2V) testing is underway.