The Grand Drive – driving without the driver
Dr. Ioannis Souflas, Research Engineer – European Big Data Laboratory at Hitachi delves into the HumanDrive project and its mammoth adventure – the Grand Drive. In this article you can uncover more about Hitachi’s role in the project, what else has been happening before the car undertakes its Grand Drive and the biggest challenge during closed road testing.
Dr Souflas will be at AutoSens in September presenting the AI/ML technology to the audience at AutoSens, tech being deployed in HumanDrive.
HumanDrive without the human
Autonomous vehicle project, the ‘HumanDrive’ is two and a half years in the making, led by Nissan’s European Technical Centre, as a part of Renault-Nissan Alliance research activities. It culminates in the most complex journey yet, the Grand Drive, attempted across the UK without driver input. The ‘HumanDrive’ project vehicle will be expected to deal with a variety of UK unique driving scenarios, including country roads, high speed roundabouts, A-Roads, Motorways in live traffic and different environmental conditions. Additionally, the vehicle will emulate a natural human driving style, providing an enhanced experience for the occupants.
To achieve this, the project will draw upon the expertise of a variety of organisations, including the tech giant Hitachi and Transport Systems Catapult, who oversaw the first UK test of a driverless vehicle in a public space in 2016. Other partners include Cranfield University, University of Leeds, HORIBA MIRA, Atkins, Aimsun Ltd, SBD Automotive and Highways England.
HumanDrive is jointly funded by government and industry. The government’s £100m Intelligent Mobility fund is administered by the Centre for Connected and Autonomous Vehicles (CCAV) and delivered by the UK’s innovation agency, Innovate UK.
Hitachi: the driving research force
Hitachi are leading the research and development of AI and communication work package. For this, Hitachi have established a new team in the UK which, together with Hitachi’s existing European Automotive Teams, designed, prototyped and tested cutting edge perception and planning solutions by employing state-of-the-art data science and machine learning tools.
Prior to the Grand Drive the HumanDrive team has been developing and testing the various functionalities of the autonomous vehicle on test tracks and in simulation. The performance of the vehicle is evaluated with respect to safety and human-likeness in collaboration with other project partners who have significant experience in autonomous vehicle testing and human factors.
One of the biggest challenges faced throughout the HumanDrive project has been the definition of “natural human like driving style”. The HumanDrive team have found that people have different preferences and there is not a single driving style that passengers prefer. However, the technology developed as part of the HumanDrive project can unlock the potential of personalised autonomous vehicles.
Watch latest videos from AutoSens on YouTube
► Panel from AutoSens Brussels: Practical experiences and future European funding opportunities
► Increasing the adoption of LiDAR through Tier 1 partnerships with AEye
► The biggest trends for solid state LiDAR with Imec
Hitachi’s test trials during the HumanDrive project revealed that people have different thresholds for considering an autonomous vehicle safe and comfortable. With this in mind the team developed technology that can be easily adapted based on customer’s needs. Data-driven automotive and infrastructure engineering is key to unlocking the full potential of autonomous vehicles and increase the safety and comfortability to levels that are acceptable to the end user.
The exact date of the Grand Drive has not been finalised yet, but it will be taking place at some point in the last quarter of 2019. The general public will be updated about the progress of the Grand Drive through social media and the official HumanDrive project website (https://humandrive.co.uk/).
Increasing public awareness of and confidence in autonomous vehicles
Building confidence in any new technology requires the end user to interact with the new product. Similarly, to increase public awareness of autonomous vehicles some level of interaction would be required with the end user. This could be achieved incrementally by using autonomous vehicles in controlled low risk environments and gradually transitioning to uncontrolled more complex situations. In parallel to that, Public Authorities can support the demystification of autonomous vehicle technology by funding educational programs, seminars and public talks.
The importance of technology acceptance at AutoSens
HumanDrive aims at using machine learning to develop natural human-like vehicle control. As part of HumanDrive, Hitachi is leading the Artificial Intelligence work package which is focusing in two fundamental challenges of autonomous vehicles, the perception and planning systems.
Interaction with rest of the autonomous vehicle technology community is a great way to discuss the advantage and disadvantages of our approach. Hitachi is committed in developing pioneering, safe and reliable technology to solve social issues (what we call “Social Innovation Business”), and this can only happen if we make sure that our technology is accepted by industry experts that can usually be found in conferences such as AutoSens.
Hear from Dr. Ioannis Souflas, Research Engineer – European Big Data Laboratory at Hitachi on “Data-Driven Perception and Planning Methodologies for Autonomous Vehicles” The presentation concentrates on the use of state-of-the-art data science and deep learning tools for representing the environment around the vehicle and imitating the human driving behaviour to improve the passenger experience of autonomous vehicles.
Tickets to AutoSens in Brussels are still available. Early bird discounts apply >>
Read latest news and updates from AutoSens
- AutoSens Award Winners Crowned ~ 14 December 20
- You bring the talent, we’ll bring the tools – with ON Semiconductor ~ 9 December 20
- AutoSens Awards 2020: Finalists Revealed ~ 2 December 20