New technologies such as AI and 5G are set to revolutionise the way we travel and commute. Advanced mobility forms and mobility services offer many benefits for society, but also introduce complex technological challenges.
While embedding individual systems into the existing mobility ecosystem is challenging already, linking the ever-growing number of systems adds further complexity and creates new risks.
Some providers have tried to introduce new systems too soon. Automated or autonomous vehicles are well-known examples, which have caused accidents and injuries. Incidents such as these are just one of the reasons why some people are now less willing to accept new mobility forms. Failures of mobility service providers to abide by data protection standards is another.
“The impending transformation of our mobility will bring profound change. Thus, gaining the trust of future users is imperative for mobility providers. Especially in new and disruptive projects, trust must be embedded as key factor from the very beginning. Only by establishing trust and ensuring the safety of complex technologies, future mobility initiatives can be rolled out comprehensively and efficiently.”
Comprehensive planning and a targeted strategy are essential for creating acceptance and trust.
Many disruptive use cases and services in the mobility ecosystem are dependent on regulators allowing autonomous vehicles.
More complex technology needs efficient management to get use cases to market quickly.
Complex regulations and resulting uncertainty are a major hurdle for both manufacturers and operators.
The mobility ecosystem is characterised by large numbers of stakeholders, all with their own interests.
AI is an essential technology for many use cases. However, ensuring safety is not a foregone matter and extended safety evidence is required compared to traditional software.
AI usage can make mobility safer, more efficient, more convenient and more environmentally friendly – providing a wide variety of benefits to mobility users.
Autonomous and connected vehicles
According to statistics by the German Federal Statistical Office, human error is responsible for 88.5% of all fatal road traffic accidents in Germany. Autonomous and connected vehicles have a huge potential to improve road safety in this context. They can also help make mobility more inclusive and more convenient.
Intelligent and connected infrastructure
Digital transport infrastructure offers many new possibilities – such as “intelligent” car parks to make parking space search more efficient, or smart road infrastructure to continuously evaluate traffic flow and incidents against set KPIs. Data recorded by such infrastructure allows traffic jams to be avoided and traffic flows optimised – for example, by implementing automatic route diversions.
Intelligent vehicle management
Intelligent systems can help make management of public transport or logistics fleets more efficient and environmentally friendly. There are also benefits for private users, such as intelligent charging systems enabling electric vehicle owners to charge their vehicles outside of peak demand – this saves money while also reducing pressure on the electricity grid.
Intelligent services
Intelligent services delivered through apps offer users a completely new and even more convenient experience. They can enable optimised multimodal transport, taking individual preferences into account.
The use cases listed here are just some of the possibilities enabled by AI.
Our expert teams are individually assembled according to the needs of each project. This allows us to tackle a wide variety of issues around trusted mobility in a targeted way – from strategic orientation and regulatory assessments right up to support for development, approval and operationalisation.
Our wide-ranging technological and regulatory expertise enables us to provide any of our clients with market-oriented support on both the strategic and operational level – eventually ensuring the success of diverse mobility projects.
We have a strong network with relevant stakeholders in the mobility ecosystem and around AI. We’re always up-to-date on the latest technological and regulatory developments, and we help to shape them through our work.
We take a holistic, end-to-end perspective when looking at mobility. This allows us to fully consider complexity and existing risks – from ideation, through development and operationalisation, all the way up to phase-out.
Technologies such as AI undoubtedly offer many new opportunities for future-oriented, user-centric and more eco-friendly mobility. But the nature of these technologies also introduces new risks. Many AI systems, for example, use machine learning, where the system develops its behaviour and decision-making logic based on historical training data. This process can lead to non-explicable and non-robust systems.
AI systems are only explainable to the extent humans can understand how a decision was made and what the outcome implies. Ensuring this level of transparency is much easier said than done, as these systems have millions or even billions of different parameters within a complex network of nodes and connections.
The concept of robustness in AI refers to a system’s ability to work properly even in unfamiliar conditions – in other words, its ability to generalise. For example, an autonomous vehicle which has only been trained using data from urban centres should still be able to drive safely in rural areas. This ability is essential because the variety of different road users, road characteristics and other factors leads to infinite different scenarios that can arise in real-world traffic. It’s impossible to consider every scenario in the development process – so even if a system has only been developed and approved for a particular operating area, it still needs to reliably know and recognise its own limits. This makes AI robustness a key factor.
“In a worst-case scenario, technical challenges related to AI explainability and robustness can cause accidents and injuries. Providers need to take every measure to minimise these risks.”
When new risks emerge due to the use of AI, it’s very important that users can trust in a safe functioning of systems. This applies in particular to mobility, where AI systems are used to obtain a high level of vehicle autonomy although developers cannot fully explain the systems behaviour.
“If users can’t understand how complex autonomous systems work, it’s even more important that they can rely on the systems being safe. Only then providers can ensure that users will accept and use their systems.”
Our study on the acceptance of autonomous vehicles in Germany, conducted at the beginning of 2023, underscores this statement: trust plays a very important role, not just for vehicle users but for all road users. Apparent reservations of the German population about the technology implies that systems must be safe and transparent. These findings can be used to develop recommendations that make it easier to successfully introduce the technology.
Learn more
The mobility ecosystem is already heavily regulated. However, the current regulatory landscape is also very dynamic, as the technologies now being introduced are new and disruptive. Therefore, existing regulations frequently show gaps or general misfits. For manufacturers and providers, this is a major source of uncertainty. At the same time, many initiatives for new regulations are gaining traction which must be observed and considered.
One of these regulatory initiatives is the EU AI Act, proposed by the European Commission, which will apply to all sectors and all forms of AI. Its requirements include risk evaluation of any AI systems by providers and regulators, whether for autonomous vehicles or intelligent apps. This will mandate some developers to meet large numbers of new requirements for new products.
“Ensuring full compliance with current regulations and anticipating upcoming ones is very complex. Tailored governance and compliance systems are therefore key to the success of your project.”
Alongside general regulations, it’s also important to consider future regulations and sector-specific requirements. For the approval and operation of autonomous vehicles in Germany, for example, you’ll need to comply with the Law on Autonomous Driving and the Autonomous Vehicle Approval and Operation Ordinance (“Autonome-Fahrzeuge-Genehmigungs-und-Betriebs-Verordnung”, or AFGBV). However, despite having a legal framework in place, manufacturers need to overcome various challenges, which are slowing down the approval of autonomous vehicles – and hence, their introduction on the market.
Our whitepaper takes a look at current challenges in getting autonomous vehicles approved and lays out practical recommendations for addressing these challenges.
Learn more
There are many different stakeholders with individual interests in the mobility ecosystem. It is essential for manufacturers and operators to understand these interests and to consider them, both in upcoming and in existing projects. While this requires additional efforts, it will pay off in the long term – ultimately, providers and other stakeholders can only solve the immense challenges of transforming mobility if they work together.
“Manufacturers and providers of advanced mobility forms and mobility services are part of a large mobility ecosystem with many different stakeholders. To ensure a successful project, it’s essential to reconcile these individual interests.”
Current safety standards for traditional systems are often insufficient to cover the complexity of AI, particularly regarding autonomous vehicles. Today’s automotive safety standards, such as ISO 21448 (“Safety of the intended functionality”) or ISO 26262 (“Road vehicles – Functional safety”), inadequately consider the risks created by AI. Since the data-driven development process creates new risks, manufacturers need to implement further AI-specific safety measures.
“AI’s unique technical characteristics require manufacturers to analyse new risks thoroughly. This makes it necessary to create additional safety artefacts throughout the AI lifecycle which specifically address the risks of a data-driven development process.”
Our LinkedIn post on autonomous driving offers guidance on this topic and explains why current automotive industry safety standards are inadequate, which aspects manufacturers need to consider for a comprehensive safety case and what AI safety artefacts might look like.
Ensuring the safety of AI systems is very challenging. As a result, manufacturers need dedicated quality control systems as well as enhanced safety cases. This involves in-depth expertise and specific methods. There is a public desire for dedicated quality control to increase trust in these new technologies, as demonstrated by the results of our study on the acceptance of autonomous driving in Germany.
To address this important concern, we founded "CertifAi" as a joint venture between PwC, DEKRA and the City of Hamburg's Innovation Fund. The plan for CertifAI is to provide support to manufacturers, helping them develop their AI systems to maturity, and getting them audited and certified. CertifAI’s services are a complement to our portfolio at PwC, making it possible to safely implement and operate various forms of AI – not just in mobility.
“Transformative change in mobility brings a wide variety of challenges. We need to tackle these challenges efficiently – and we can only do this if we work together in the ecosystem. Manufacturers and operators will profit from reliable partners helping them to consider trust from the very beginning, which is a key factor to transforming mobility.”
Although Germany already has a legal framework for approval and operation of autonomous vehicles (SAE level 4), manufacturers are still struggling to actually get their vehicles approved. Our whitepaper sheds a light on this topic, examines current challenges in the approval process and provides recommendations on addressing them.
“Transformative change in mobility brings a wide variety of challenges. We need to tackle these challenges efficiently – and we can only do this if we work together in the ecosystem. Manufacturers and operators will profit from reliable partners helping them to consider trust from the very beginning, which is a key factor to transforming mobility.”
Hendrik Reese, Partner, PwC Germany
