ITS UK has a Memorandum of Understanding in place with the Automotive Electronics Innovations Network (AESIN), one of six networks that form part of TechWorks, allowing mutual participation in events and dissemination of information. I had the pleasure of attending the AESIN plenary conference on the 19 June at the British Motor Museum at Gaydon.

This is a well-attended event with a cross-section of industry representation. Part of the reason for attending is to continue the dialogue of how the vehicle and the systems used by network operators need to be more connected if we want to realise real benefits in efficiency, journey times and safety. However, before we determine if the view is held in a similar level of esteem as it is on the ITS UK side, here is a summary of the event’s proceedings.

Overview

The automotive electronics sector is changing and growing at a rapid pace. For the UK, this is a £67bn per annum industry employing over 1m people. There is considerably more requirement for electronics in the vehicle that there was historically to meet changes in propulsion, electrification and connectivity. Buyers are demanding more personalisation from their vehicles, such that they fit in with their lifestyles, and automotive companies believe there is value in selling services to customers to enhance their ownership experience. This is requiring the industry to evolve and try new business models, which is blurring the lines between the traditional tiers that people had been familiar with historically. It presents both opportunity and threat across the supply chain, but most see the former.

Rationalisation of the compute platforms on vehicles

One of the big topics in the automotive sector, which is potentially of significant interest to those in the ITS industry, is the rationalisation of the compute platforms. Currently vehicles can have over 70 distributed processors around the vehicle, almost all bespoke to their task and requiring considerable effort to define their requirements and integrate them. Moving to a centralised compute platform opens the door to the Software Defined Vehicle (SDV), which would allow updates to the vehicles functionality over time. What might take a few lines of code to create a feature to prevent a door opening if it was obstructed would require 50 software and 10 hardware modules using existing architectures.

This potentially opens the door to being able to easily create integrations with ITS systems to enhance data exchange and functionality that benefit both the vehicle occupants, but also those sharing the network with that vehicle. Importantly, it would enable continual development and deployment of new or enhanced features, with relative ease compared with the approach to vehicle architecture.

Testing and verification of advanced driver assistance systems

As vehicles become more connected and automated, number of additional technologies and software code mean they are also increasingly complex, making them more challenging to test and certify for use on public roads. A number of strategies are being considered across the development lifecycle to address these challenges, including simulations, hardware in the loop, vehicle in the loop, parallel simulation and mixed reality testing. Underpinning a lot of the testing is having models with an appropriate level of fidelity to provide confidence that behaviours exhibited in the model will be reflected in real-life.

Digital-twins has been the buzz phase to describe this and we have seen an increasing case to justify expenditure on these models to support more than just vehicle testing, but also for infrastructure management and user behavioural studies (along with many other use cases). For the ITS sector, the Digital Traffic Regulation Orders (DTRO) provides a great example of the types of input one of these models may use to demonstrate a vehicle would behave as expected.

Recall numbers are growing and most are related to software

Despite the improvements in testing, it is still trailing the technological advancements on vehicles. With software on vehicles containing more lines of code than any other equivalent (such as aircraft or space rockets) it is not a surprise that the number of software and electronics related recalls have increased from circa 5% in 2011 to 65% in 2022. 20% of all recalls are software patches and one OEM spent over £2.1m in postage alone to notify owners their vehicle for a single recall.

This will drive improvements in approaches to the development of these systems, but combined with the elongated development timescales in bringing products to market, is forcing some of the change highlighted earlier. This is further compounded by safety related software costing five times as much to test and certify. There will be more pooling of resources and standardisation of platforms, greater reliance on third-party products to provide enhancements, such as Apple CarPlay Ultra. This may also drive the debate about the line between connected features and vehicle automation, can more be done by the network operator to supplement the vehicle?

Cyber-attacks are becoming more commonplace

From 2019 to 2023 there has been a 50% increase in the number of cyber-attacks on vehicles. UN155 mandates the security of in-vehicle systems, but their complexity and bespoke nature makes this highly difficult to maintain throughout the vehicle lifecycle. Like many systems we use day to day, they are frequently patched to remove vulnerabilities or address new attack vectors. Noting the numbers recalled for software issues, this will further compound the requirement for regular updates. This is far from ideal for the customer, but may also present issues for systems connected to these vehicles, including those employed by transport authorities. Standardisation of interfaces and security requirements will be essential to achieving trustworthy solutions that enable the connectivity to deliver the efficiency and safety benefits that these systems can achieve. 

Views from the floor

Some of the interesting takeaways from the day were the views and impressions of the speakers and delegates:

• Automated driving won’t solve our road safety issues;
• Vehicle manufacturers still aren’t in the mindset of connectivity between the vehicle and infrastructure;
• OEMs are hoping to capitalise on services they can sell to customers with the expected revenue per manufacturer ranging from $1bn – $26bn;
• Brand loyalty is in decline, particularly in the UK, where connectivity features are customer priorities and we are seen as a prime target for new entrants to the vehicle market;
• The UK has the talent, innovation and track record to deliver high quality products and services to the global automotive sector; and
• The industry, like many others, needs more diversity.

Get involved

Please do get in touch with myself or Manish Kumar if you want to have greater involvement in AESIN or if you’d like to request or propose topics of discussion for our next CAV Forum meeting. You can also contact us at contact@its-uk.org