Skip to main content

AUTOmated driving Progressed by Internet Of Things

Periodic Reporting for period 2 - AUTOPILOT (AUTOmated driving Progressed by Internet Of Things)

Reporting period: 2018-07-01 to 2020-02-29

The automotive industry is going through a transformation: On one hand, autonomous vehicles are quickly gaining traction and will soon take over the majority share of the automotive market in the next two decades. Analysists are of the view that autonomous driving cars will make up about 75% of the global light-duty vehicle market by 2035 with about 95.4 million units by then. On the other hand, already in 2020, one in five vehicles is expected to have some sort of wireless network connection, accounting for more than a quarter of a billion cars on global roads. The connected car market is gaining a boost from the advances in the Internet-of-Things (IoT).
In the last decade, the research community has developed a large number of IoT technologies, making it possible to deploy IoT-based solutions and provide new services. The developments include techniques for the identification and discovery of internet connected devices and non-connected physical things, technologies for modelling data and services, IoT software engineering tools, schemes for safeguarding security/privacy, as well as infrastructures for deploying and operating IoT services within cloud computing infrastructures. The growing connected car community is a prime group for innovating on the basis of IoT concepts and technologies. With the market for automotive IoT and connected cars projected to reach a staggering $133 billion by 2024, it is to be expected that cars will be a “major element” of the expanding Internet of Things.
Automated vehicles today rely largely on on-board sensors (LiDAR, radar, cameras, etc.) to detect the environment and make reliable decisions. However, the possibility of interconnecting surrounding sensors (cameras, traffic light radars, road sensors) to reliably exchange complementary data could lead to new and improved ways of designing automated vehicle systems with reduced implementation costs.Connected cars and overall ITS solutions need to become horizontally integrated with IoT platforms/systems in order to benefit from self-configuration, device discovery, IoT-based services, data filtering, brokering and shared semantic world models of their environment. These communities, however, currently face some difficulties when it comes to taking advantage of IoT technologies. This is mostly due to the lack of open standardized and easy-to-use APIs for accessing IoT technologies, but also due to the lack of essential interoperability between ITS systems and IoT platforms.
AUTOPILOT took on the mission to address this important challenge. The large-scale pilot set out to develop and validate sustainable solutions for automated cars, within the Internet-of-Things. The AUTOPILOT project’s aim was to use the possibilities offered by IoT for automated driving (AD), and at the same time to also make data from autonomous cars available to the Internet-of-Things to enrich it further.
AUTOPILOT’s goal was to bring together knowledge and technology from the automotive and the IoT value chains in order to develop IoT-architectures and platforms and bring automated driving towards a new dimension.
AUTOPILOT partners developed and integrated the IoT platforms in vehicles and other devices while developing and adapting AD functions for the IoT-progressed AD uses cases. The AUTOPILOT open IoT service platform, a federation of several vendor-IoT platforms provided by the project partners and interconnected through an open oneM2M standard IoT platform was set up.
AUTOPILOT partners developed a common approach to evaluation based on the FESTA methodology. The technical evaluation methodology defined requirements relating to data log formats including vehicle data, environmental detection, communication (IoT messages), application and data management logging. The methodology for the Business impact evaluation identified the parameters for using the Cost-Benefit Analysis, SWOT, Multi-Actor Multi-Criteria Analysis tools. The quality of life evaluation methodology used the impact assessment framework developed by the Trilateral (EU-Japan-US) Working Group on Automation in Road Transportation. Requirements for the evaluation methodologies user acceptance assessment and legal issues were also established.
In the second phase of the project, the focus was on testing and evaluating AUTOPILOT’s IoT enabled AD use cases.
AUTOPILOT’s IoT enabled automated driving cars were tested with various services in real conditions at six permanent large scale pilot sites. Four modes were trialled:
1) Automated Valet Parking (AVP) in Brainport, Tampere, Versailles and Vigo
2) Urban Driving applications in Brainport, Livorno, Versailles, Vigo, Tampere and Daejeon (Korea)
3) Highway Pilot in Brainport and Livorno
4) Platooning in Brainport and Versailles
The test results at pilot sites were put to multi-criteria evaluations (technical, business, quality of life, user acceptance and legal issues) of the IoT impact on pushing the level of automated driving.
AUTOPLIOT's use cases and the vehicle IoT platforms highlight the role of role of the IoT sensor data and the possibility to use the vehicle as an IoT device, which increase the amount of driving-relevant data to be collected and thus to improve the driving environment perception significantly, easing the way to implementing high-level of automation (L4).
Working on the needs to implement interoperable vehicle and cloud IoT platforms, together with heterogenic IoT sensor devices, triggered the success of defining a common data model. This common data model development has been carried out jointly with the SENSORIS platform, responsible to provide standards for the vehicle to cloud data. Relevant standards have also been presented to OneM2M. AUTOPILOT Platooning and Automated Valet Parking use cases are used as references in the ETSI Technical Report TR 103 508 (https://www.etsi.org/deliver/etsi_tr/103500_103599/103508/01.01.01_60/tr_103508v010101p.pdf) “SAREF extension investigation: Requirements for Automotive” (note: SAREF: Smart Appliances REFerence ontology). SAREF is an essential standardisation work to ensure intra-domain and x-domain interoperability. AUTOPILOT Open Data will help standardisation organisation and researchers to define common data model for the automotive domain. In the context of the evaluation, the FESTA methodology has been enhanced for including the IoT data.
By demonstrating the technical innovation and business case around new mobility services using automated driving helped by IoT, the AUTOPILOT project has highlighted a number of advantages for users in terms of safer and more comfortable journeys. With IoT expanding the perception environment of the vehicles can driver longer and with fewer sudden manoeuvres in autonomous mode. Smoother riders also result in less pollutants and harmful emissions. Through IoT, users can avoid crowded routes and also enjoy the comforts of shared mobility services, such as car sharing. Besides specific benefits for individual drivers or fleet operators (platooning, car rebalancing), AUTOPILOT’s results will also be advantageous for Mobility-as-a-Service and Smart Mobility in general. One of the biggest contributions of AUTOPILOT is validating the role of the automotive sector in the digital marketplace. AUTOPILOT’s use cases, especially the highway pilot, can be used to further the EC’s agenda on cross-border societal challenges as prioritised in the CEF2 agenda.
AUTOPILOT - Federated IoT Architecture
AUTOPILOT - Automated driving progressed by IoT - How does it work
AUTOPILOT - Final event
AUTOPILOT - Project logo