WARA-CAT: State-of-the-art research arena for Collaborative and Autonomous Transport research
The Collaborative and Autonomous Transport (CAT) research arena aims at providing leading edge research facilities and host active research collaborations for testing, exploring and demonstrating highly automated driving and C-ITS (Collaborative Intelligent Transport Systems). Certain emphasis is put on city traffic, with the goals of increasing the level of automated driving support, improving traffic efficiency, and maximizing safety. In order to enable a complete system setup the scope reaches far beyond individual autonomous vehicles, including vehicle-to-vehicle communication, advanced communication with infrastructure as well as different forms of cloud services. In addition, research and demonstrations about methods, tools, and processes for working with an endlessly growing amount of data from and about autonomous systems is crucial.
In order to gather groups of researchers around specific and realistic traffic scenarios a handful of use cases have been defined. In this way, in depth research in more narrow research areas (e.g positioning, object detection, edge cloud processing) can be combined and integrated with other fields of research, thus enabling a complete system setup. The use cases identified this far include Roundabout, Intersection, Highway, Geo-fencing, Bus stop, and Verification of autonomous systems. Remote Driving is being discussed as an additional use case. Moreover, by combining different use cases, more complex traffic environments can be created. For example, several intersections and roundabouts can be set up to create a city-like traffic environment.
By developing state of the art resources and equipment for exploring, testing, verifying and demonstrating research, WARA-CAT provides well-prepared research platforms letting researchers focus on their research.
The research platforms include several test vehicles (research concept vehicles, passenger cars, trucks, trailers) equipped with various sensors (e.g. radar, lidar, cameras etc), high precision GPS systems, communication equipment, logging equipment etc. Software platforms (middleware) for autonomous vehicles are provided where researchers can easily implement algorithms, connect new sensors etc.
The test vehicles can be driven on public roads as well as on dedicated test tracks. One such test track is AstaZero, located about one hour drive east of Gothenburg, Sweden. Also other test tracks located closer to Stockholm and Linköping can be used. At these test tracks, different traffic environments (e.g. roundabout) can be built up when needed, thus supporting different use case related tests.
At the moment testing of autonomous functions (including vehicle actuation) can be done at test tracks only, while public roads are mainly used for collecting real-traffic data. However, work is on-going to apply for special permits to also enable testing of autonomous functions on public roads.
For efficient communication between vehicles as well as between vehicles and infrastructure a dedicated pre-5G end-to-end Proof of Concept network has been built up at AstaZero. To start with, the network consists of the latest LTE technology with some concept software such as Network Slicing, Local Breakout, and a Distributed Mobile Edge Cloud to minimize latency. The network is dedicated to this purpose and can be optimized depending on use cases, and in this way provide better characteristics than any commercial network with the same features. The network is evolving towards full 5G when these components become available and use cases so require.
Furthermore, an environment (hardware and software) for distributed mobile edge and centralized cloud computing is being set up, and the software needs are being looked into from the perspective of for example analytics, machine learning, and certain time-critical functions.
During the fall of 2017 WARA-CAT will be focused on supporting groups of researchers engaged in the WASP Project Course. The following five project course themes are connected to WARA-CAT.
- Coordination of vehicles over C-V2X
- Edge Cloud and IoT Platform
- Offline SLAM for automotive
- Reinforcement learning for autonomous vehicles
- Discrete Control of Research Concept Vehicle
Soon researchers will get the opportunity to sign up for the different use cases. The first tests and demonstrations in the use case environments will be developed during spring 2018 and the plan is to arrange at least two larger demo events open to the WASP community during 2018.
WARA-CAT is managed by the WARA-CAT Core Team that includes representatives from academia (researchers and research labs) as well as industry. This far, the following organizations are represented, and additional partners are welcome to join.
- Linköping University
- Lund University
- Chalmers University of Technology
- KTH Royal Institute of Technology
- Ericsson AB
- Volvo AB
- Volvo Car Corporation
The possibility for exploring, testing and demonstrating research in real-traffic like conditions is crucial for building knowledge and experience as well as improving the credibility for the research. If you want to learn more about WARA-CAT and how to access and utilize the resources described here, don’t hesitate to contact WARA-CAT Manager Fredrik von Corswant, frecor at chalmers.se