Connected and cooperative autonomous systems projects
CCAS Projects
Projects
To find out more about projects from the Connected and Co-operative Autonomous Systems (CCAS) group, email wmgbusiness@warwick.ac.uk
Adaptive Autonomous Towing for MObile robots (AATOM)
The transport of equipment, material, parts and waste within a controlled environment is a common requirement across a wide array of sectors. While the environment and payload may change, the underlying challenge remains the same. This task is often labour-intensive, making automation an ideal solution to enhance efficiency. Towing offers an effective way to extend the carry capacity of vehicles and is a common practice across industries such as agriculture, manufacturing, construction, and aviation. In these sectors, tractor-trailer systems are frequently employed for transporting goods.
The challenge to automate this process, however, is that a trailer does not necessarily follow the same path that the vehicle does. The figure below illustrates this problem by highlighting (in orange) the areas where a small trailer deviates from the vehicle’s path. This problem is propagated when the size of the trailer is increased or when the system includes multiple trailers.
Moreover, towing enables additional benefits that could not be realised with a vehicle alone:
Increased modularity by allowing one vehicle to tow many trailers with different usages
Increased utilisation by allowing one vehicle to move multiple different trailers meaning that the vehicle does not need to wait for the trailers to be loaded or unloaded
Minimised costs by being able to minimise the number of vehicles required
Scalable solution by allowing for multiple trailers to be towed by 1 single vehicle
We are currently developing this technology for use within Agritech however this is the most challenging environment due to the weather and terrain involved. It will be possible to transfer the same technology and learnings to other simpler environments.
Carma
Project Description The project focuses on unleashing the power of Edge and Cloud Computing, using a “connected roundabout” at the University of Warwick’s main campus.
The Cloud Assisted Real-time Methods for Autonomy (CARMA) project, is part of the £11m TASCC programme funded by the Engineering and Physical Sciences Research Council (EPSRC) and Jaguar Land Rover. Supported by WMG’s Centre High Value Manufacturing Catapult, the CARMA project was established with the intention to create secure and resilient cloud-based platforms to enable safe and robust semi-autonomous functions on future cars in the short term, and with the vision of achieving fully autonomous vehicles.
The testing infrastructure was supported by the Midlands Future Mobility project, Innovate UK and the Centre for Connected and Autonomous Vehicles of the UK Government. The Open Innovation Platform research vehicle used within the project was developed with support from the High Value Manufacturing Catapult and used alongside the CARMA research vehicle provided by JLR.
Our role Intelligent vehicles researchers installed eight infrastructure cameras, as off-board sensors, at the roundabout to monitor the environment and stream video to a base station called ‘Edge’. Using two-way communication, the Edge processes its own live information with information received from nearby connected vehicles. This processed data containing object, traffic, road layout and lane availability information is broadcast and received by the vehicles. With the enhanced perception provided by this information, we can use our autonomous research vehicles to showcase how these technologies work in practice, demonstrating how autonomous vehicles can make more intelligent and robust decisions.
Achievements
Industrial Impact
Over ten follow-on projects have been inspired by CARMA, as technological, regulatory and business challenges are to be overcome before the concept can be deployed on a mass scale. Through the project 23 early career scientists and engineers were trained and three patents were granted along with 40 publications on the work carried out.
Core partners JLR, TRL, University of Surrey, EPSRC, TASCC, UKRI
Deployment of Cooperative Autonomous Vehicles in Agritech
Project Description The UK horticulture industry is facing several challenges including, but not limited to:
Labour Shortages due to increasingly aging workforce and difficulty finding labour in the UK.
Lower crop quality and increased waste. Implementing quality control and monitoring systems is expensive, as a result farmers are more comfortable wasting a small percentage of crops.
The introduction of autonomy in horticulture would have the potential to resolve these issues in addition to increasing overall process efficiencies, resulting in massively increased crop yield and product quality.
The project leverages capabilities from both Intelligent Vehicles and the Warwick AgritechLink opens in a new window group, addressing some of the current challenges in the UK horticulture sector by developing a crop monitoring use case for an autonomous mobile robot (AMR).
Our role The Intelligent Vehicles group were primarily responsible for the development of the robot’s autonomous navigation system. This included the localisation, perception, path planning, and control algorithms for the vehicle. The integration of an advanced sensor suite into the Catapult funded mobile robot allowed the robot to sense, process, and navigate its environment in real time. To enable non-technical audiences, such as farmers, to operate the robot, a user-friendly dashboard was created, allowing the user to seamlessly interface with and control the robot over Wi-Fi. Additionally, two supplementary use cases were developed and trialled during the project; utilising infrastructure sensing to provide additional information to the vehicle of other greenhouse occupants, thereby enhancing safety and efficiency, and a Wi-Fi button that could be used to call the robot to a specific location or have it follow the button holder.
Achievements An industry engagement event was hosted, showcasing the project developments, and outlining the potential impacts of this technology in the UK horticultural industry.
Industrial Impact The first year of the project laid the groundwork for future work in the area of Agri-Tech. In recognition of the achievements in the first year of the project and the importance of this work for the wider industry, Warwick have formed an Agri-Tech department; addressing issues such as labour shortages, food insecurity and loss of biodiversity. The AATOM (Autonomous Adaptable Towing for Mobile Robots) project is a direct continuation of the work carried out in the first year, investigating further use cases for autonomous mobile robots in agriculture. Moreover, the technologies developed by WMG’s Intelligent Vehicles group, primarily for autonomous vehicle development, can be scaled to suit a multitude of industries and use cases, such as horticulture, manufacturing, and construction.
Hi-Drive (https://www.hi-drive.eu/) is a 60 million Euro EU flagship project with 53 partners across 13 countries. Started in July 2021, Hi-Drive project aims to advance the European state-of-the-art of automated driving from SAE L3 ‘Conditional Automation’ further up towards ‘High Automation’ by demonstrating in large-scale trials the robustness and reliability of Connected Automated Driving Functions (CADFs). We focus on testing and evaluating variety of functionalities, from motorway chauffeur to urban chauffeur, explored in diverse scenarios with heterogeneous driving cultures across Europe.
The complete set of conditions that allow safe operation of an Automated Vehicle (AV) is defined as the Operational Design Domain (ODD). Today, even within the parameters of the ODD, the AV encounters a number of challenges that require the driver to switch back to manual driving from automated driving mode. This situation compromises the overall travel experience and is unacceptable for a marketable vehicle.
Hi-Drive strives to extend the ODD and reduce the frequency of takeover requests by selecting and implementing technology enablers leading to highly capable CADFs. Passenger cars and trucks will demonstrate CADFs in a large set of traffic environments on motorways, in cities and cross-border scenarios, with a specific attention to demanding, error-causing conditions.
Connected Automated Vehicles (CAVs) equipped with robust and reliable technology enablers bridge the ODD gaps. The removal of fragmentation in the ODD paves the way towards a gradual transition from a conditional operation towards higher levels of automated driving. The result is a safe and efficient automated road transport system, in which CAVs operate for longer periods and interoperability is assured across borders and brands.
Our role
Hi-Drive removes the fragmentation in the ODD through robust and reliable technology enablers implemented into the vehicle and infrastructure and demonstrates the defragmentation through well-defined operations.
WMG’s role is to complete two operations utilising nodes (vehicle and infrastructure) developed through previous projects. In addition, WMG leads the V2X communication enabler group on the Hi-Drive project and spearheaded the research on Cooperative autonomy.
The two operations designed to fulfil the Hi-Drive project target are:
1-Urban Operation:
Improving the automated driving function at urban environments that includes a roundabout. This operation has three use cases and utilizing two enablers besides other infrastructures.
The two enablers used are:
a-V2N for Cooperative Sensing
This enabler implements cooperative sensing between the EGO vehicle and the Edge/infrastructure using V2X communication over ITS-G5. The goal is to improve the understanding of the road traffic conditions at the EGO vehicle with the comprehensive understanding of the environment obtained by the Edge, while the EGO vehicle travels across the (unsignalized) roundabout.
b-3D Cooperative Object detection
The enabler implements a cooperative 3D object detection model using deep neural networks. The model utilises sensory data from multiple spatially diverse infrastructure sources to increase the robustness of detection. The enabler has two main components: (I) a 3D object detection module that localises and classifies objects of interest in the environment, and (II) a data fusion module implemented at the infrastructure (or Edge) PC that combines data collected from two infrastructure LIDAR sensors.
The infrastructure sensors are linked to the central fusion system through wired/wireless high-speed data links. The performance will be evaluated at different detection fusion levels: early fusion where raw point cloud data from the two LIDAR sensors is fused and late (object level) fusion. In both cases, the 3D object detection and classification leverage a neural network with the CenterPoint architecture.
2- Motorway Operation:
The purpose of this operation is to collect real-world data for evaluation of the associated enabler (i.e. Learning based Vehicle Intention and Trajectory Prediction Enabler) on motorway merging scenario. The evaluation has been carried out in WMG’s lab and it aims to first measure the performance of the enabler on real-world vehicular data and then understand the benefits of the enabler for a simulated ADF designed for motorway merging.
Achievements
Enhancing the AD robustness in challenging driving scenarios using the enablers, where our developed Automated Driving Functions (ADF) algorithms have been applied and tested in real-time under real-world traffic (complex junctions such as roundabout) at the University of Warwick campus. where the overall tests carried out were about 1121 test runs.
The testing results proved the potential of the cooperative perception technology enabler to enhance the safety, comfort, and robustness of ADFs in demanding driving conditions.
Industrial Impact
Our work highlighted the positive impact of cooperative autonomy technology, where the highly capable connected automated driving Functions (CADFs) offers robust and reliable automated driving.
This project demonstrates CADFs within Passenger cars and trucks in a large set of traffic environments on motorways, cities and cross-border scenarios, with specific attention to demanding, error-causing conditions.
Core partners
The project is supported by a consortium of 53 partners, where the lead partner is VW.
Follow the link below to find out more about the list of partners:
The European research project L3Pilot tested the viability of automated driving as a safe and efficient means of transportation on public roads. It focused on large-scale piloting of SAE Level 3 functions, with additional assessment of some Level 4 functions. The functionality of the systems was exposed to variable conditions with 1,000 drivers and 100 cars across ten European countries, including cross-border routes.
Our role
WMG at the University of Warwick implemented a proof-of-concept demonstration showing how cooperative perception can support a vehicle to navigate through a T-Junction while its own view of the junction is obscured. We installed a sensor suit including cameras and Lidars on the vehicle as well as infrastructure and connected them with wireless communication technology ITS-G5/DSRC, which enables the sharing of information between them, including identified objects and their location. With cooperative perception the vehicle was able to build up a comprehensive understanding about what is around it in the environment to then make safe, efficient, and informed decisions in comprehensive road junctions.
Achievements Cooperative perception helped extend the operational design domains of automated driving system to more challenging driving scenarios for Level 3 autonomy and beyond. It also plays a key role in enabling safety, comfort and efficiency of automated driving in complex road segments.
Industrial Impact
L3Pilot: L3Pilot Within L3Pilot 34 partners from industry and academics tackled this challenge and put more than 70 fully equipped prototype cars on public roads all across Europe, including cross-border activities. More than 750 test participants drove 400,000 kilometres, half of them in Automated Driving mode. L3Pilot has clearly shown how fragmented the Operational Design Domains (ODDs) of the functions still are.
Core partners 34 partners, among them OEMs, suppliers, research, SMEs, insurers, authorities and user groups, from 12 countries: Austria, Belgium, France, Finland, Germany, Greece, Italy, the Netherlands, Norway, Sweden, Switzerland, United Kingdom View full PDF here.
Scale
Project Description
Solihull and Coventry Automated Links Evolution (SCALE) is a Centre for Connected and Autonomous Vehicles (CCAV) funded project. “SCALE” will deploy three self-driving vehicles on a route from Birmingham International Rail Station, through the National Exhibition Centre and into Birmingham Business Park.
SCALE, led by Solihull Metropolitan Borough Council, brings together a highly skilled and experienced cohort of sector specialists from across the region. Using the innovative and intelligent Ohmio UK LTD “Lift” vehicles, the project will see vehicles deployed at the NEC in late 2024 and in to 2025.
The trial will last 3 months and is aiming to understand how automated vehicles can address some of the transport challenges of the future. Working closely with Transport for West Midlands (Part of West Midlands Combined Authority), an industry-leading business case will be developed to understand how self-driving vehicles can become a commercially viable part of our future transport.
Critical to the project are IPG automotive and d(RISK) who use simulation testing to understand how self-driving vehicles behave in the real world. Exploring “behaviours” and “edge-cases”, IPG and d(RISK) are helping to reduce real-world risk; understanding vehicle and situational behaviour over thousands of AI-supported virtual scenarios.
Our role
Warwick Manufacturing group, with its proven track record of successful collaboration between academia and the public/private sectors, will be working closely with Ohmio UK LTD to develop the pre-deployment safety case, development and analysis of the testing plan and independent academic “assurance” of testing, safety, and the Lift vehicle.
Industry Impact
This project will lead to a greater understanding of the real-world commercial justification and viability of self-driving vehicles. Working with Direct Line Insurance, it seeks to understand the insurability and repairability of these types of vehicles as we being to see them appear on our roads. With constant evolving government legislation and guidance, WMG and Coventry University, will be core to providing insight and comment to future legislation – using SCALE as a key project from which to learn. SCALE will examine current understanding of what infrastructure investments our local government partners will need to make. Self-Driving vehicles will require V2X (Vehicle to Everything) technologies that should support not just communication to infrastructure, but also the requirements of remote monitoring and tele-assist/operations. Our learnings will feed into infrastructure policy for decades to come as we move closer to smart cities.
Key partners
Solihull MBC, Coventry University, Ohmio UK LTD, CCAV, Department of Transport, West Midlands combined authority.
