Telematics Wire talks with Charles Smith, CEO and founder of Collision Management Systems (CMS). Excerpts of the interview:
What is driving data explosion in connected vehicle?
The first wave of mobile technology radically changed the way we move around our cities and beyond. By assimilating masses of data points, on vehicle position and relative road speed, it created new ways of managing fleets of vehicles and enabled real-time navigation. Today, the humble vehicle of old, with its mechanical controls and switches, has been transformed with embedded and aftermarket sensors and on-board computers into a hub capable of delivering the next wave of mobile technology. There are now over 330 million vehicles, connected via telematics or by in-vehicle apps[i], each with more processing power than put a man on the moon.
The current data explosion is being driven by sophisticated ADAS systems, the precursor to fully autonomous vehicles, which generate masses of new data points and utilise real-time video. Commercially available aftermarket systems are also sending information on vehicle faults, lane departure, driver fatigue, tailgating, proximity to other vehicles, etc. According to McKinsey, connected cars create up to 25 gigabytes of data per hour, more than a months worth of 24-hour music streaming. Looking to the future, fully automated vehicles will take this even further. They will have to gather enormous quantities of data from a huge variety of sources in order to think for themselves; to function and respond like a ‘human brain’; and make complex driving decisions in real-time. There is then the vast amounts of data that will be needed for third-parties to review issues and incidents should anything go wrong. Multiply this with the millions of vehicles on the roads and it’s easy to see how data volumes will be pushed of-the-scale.
Will connected vehicle data assist in the growth of Insurance telematics sector?
Young driver insurance has been the bedrock of insurance telematics for several years now. The reality, however, is that while there are many new drivers taking out telematics policies, people are going into mainstream insurance just as fast so total connections are plateauing. In addition, the cost of professional or even self-fit telematics makes the economics unsustainable for insurance companies as they fight to stay profitable against rising claim costs. Yet CMS believes that connected vehicles can deliver new growth for this sector. With lower cost embedded OEM devices, the economic challenge falls away. Customers will be able to link their data to their insurance policy seamlessly and their previous driving history can be unlocked to help them get a better price. All of this depends on data quality. At present, retro-fit telematics devices are rich in data, but embedded devices are often simpler which can affect data quality.
Insurers will be reliant on motor manufacturers for data sourcing, with no guarantees on what data will be available, in what format and, importantly, at what cost. At present many motor manufacturers haven’t worked out their commercial models for data-sharing.
What can be the other use cases “Data” in the automotive sector?
For automotive OEMs, there’s huge value in having real-time access to connected vehicle data to help control the customer experience around a breakdown or accident. Fast response and recovery, with rapid turnaround or replacement vehicles, can all drive revenue in the OEM’s own dealer/repair network. Sending data ‘over-the-air’, opens up two-way data communication between the OEM and the vehicle to springboard new applications. For example, diagnosing and adjusting tuning and ratio settings in real-life driving situations to improve engine and gearbox lifetime. OEMs are even using the data to analyse exactly how and when drivers are using specific features, such as sunroofs, helping them to respond to the voice of the customer more intelligently. When autonomous vehicles eventually become the norm, there will no longer be a ‘driver’ experience only a ‘passenger’ one. Current connected car services help us understand how passengers interact with vehicles and provide valuable insight on the chargeable services that could be delivered to passengers in autonomous pods in the future. Organisations such as Google and Facebook, have already shown how mobile can be used as a platform to monetise location-based services. Similarly, the growth of in-car concierge and infotainment services will make pod-users the key consumers of new datasets.
Discuss the challenges of Data Protection and Cybersecurity in automotive space and how they can be negotiated?
Recent issues with social platforms, advertising, and data-harvesting, have highlighted just how little people know about their personal data – what’s being collected, where it’s going, and how it’s being used. The introduction of GDPR is helping to address this balance, but there is a general consensus across industries that data-services need to be more transparent. Cars are already connected wirelessly to a myriad of third-parties through diagnostics, infotainment, location-based services, and telematics, which is starting to push data protection up the automotive agenda. Indeed, for our own global clients, such as Swiss Re, CMS must ensure any vehicle data is processed and the country of origin and in accordance with all relevant local and international guidelines.
It’s easy to envisage a scenario in the future, where customers request their vehicle-based data to be centrally stored, signing up to a third-party like Google, to consolidate and manage it on their behalf. As aftermarket devices become more embedded and cars become increasingly connected to the wider IoT ecosystem, the potential attack surface will become much larger, resulting in new risks and potential threats to personal safety, security and privacy. One way to simplify the challenge, while still allowing for rapid innovation, is for OEMs to develop a secure perimeter layer that acts as a ‘security fence’ for all internal systems, rather trying to lock-down individual modules. Secure in-car gateways could then ensure that any transmission/ communication was controlled using various forms of authentication. OEMs, tier-one suppliers and IT service providers will have to work together to understand vulnerabilities and develop
OEMs, tier-one suppliers and IT service providers will have to work together to understand vulnerabilities and develop cybersecurity solutions to keep car-data secure. Regulation has a key part to play in this. In 2017, the UK government issued ‘The Key Principles of Vehicle Cyber Security for Connected and Automated Vehicles’, which offers some guidelines for maintaining the security of wirelessly connected cars.
What are your views on the increased use of Analytics in the automotive sector, where do you see it going?
For the past five years, CMS has been working to make data actionable for telematics companies and insurers. During this time, we have seen the emphasis shift from historical data analysis, towards real-time analytics and alerting – particularly on driver risk.
Applications which predict events, require significantly more data – tens of billions of data points, rather than hundreds of millions. The glaring issue is that there often isn’t enough raw data to train the analytic models. In some instances, in-device filtering limits the information available so impacts its usefulness. Instead, of trying to shrink data, the emphasis should be on developing scalable, data aggregation and analytic solutions that invisibly and seamlessly automate the handling of large-scale data while retaining its richness. Delivering actionable information while minimising the data burden on people and services. Increasingly, analytics is giving us a better understanding of the driver, their behavior, their interaction with the vehicle, their surroundings and other road users. This information will form the foundation to build more personalised and targeted car-related services and safer and more enjoyable user experiences. Analytics also unlocks the value pool inherent in data, so it can be monetised by third parties, as long as the data-use is authorised in advance with the car owner/driver. For example, geo-linked advertising, personalised to the driver on in-car infotainment systems as piloted recently by General Motors. Aggregating analytics across millions of vehicles will also deliver new insight and identify trends that can be used to make roads safer, cities smarter, vehicles more environmentally-friendly and traffic more manageable.
Comment for the need for “dynamic collaboration” in automotive ecosystem.
The era of big technology companies delivering the whole technology stack is declining. Instead, they are becoming the glue that sticks all the players together. This is most apparent with Amazon Web Services and Microsoft Azure. Automotive business models are also changing and so is the technology supporting it. This requires brand new solutions to be created, not just tweaks or upgrades,
making the pace of change very high. To accommodate this, companies are specialising and co-operating even if they compete (co-opetition), driven largely by the increasing complexity and data volume required to solve certain problems using machine learning. If dynamic collaboration works, then the ability for customers to move or make their data available increases substantially, ultimately underpinning new products and business models. At the moment this is fine as this new market is growing so fast and there is room for everyone. It will become more interesting when growth starts to plateau, and companies start to focus on the new data revenue opportunities. Realising that open sharing may not be good for their short-term revenue growth (even though closing it down will harm them in the long term). Over the next few years, we will see the development of more symbiotic relationships between auto, tech and wireless players as well as governments and insurers, with the aim of making ‘mobility’ smarter, safer and more efficient. At CMS, we are already actively involved in several collaborative projects such as Transport Systems Catapult and the Silverstone Technology Cluster. We are also partnering with multiple large and small companies to deliver solutions that would be impossible alone, creating embedded end-to-end, data/telematics/connectivity propositions that improve and add value to every element in the chain.
How is Cloud empowering automotive companies to expand business agility and capability?
Cloud has enabled small companies to access the same infrastructure that used to be reserved for the IBM’s of the world. It also gives large organisations such as motor manufacturers the opportunity to work with smaller more innovative organisations,
who can use the cloud to deliver new types of specialist applications and services, faster, more nimbly, and with less resource than if they tried to do it themselves. The cloud is also allowing OEMs and insurers to exploit SaaS services, to enhance the accuracy and functionality of their telematics solutions. At CMS, our software can be licensed and supplied over the cloud to embed new capability for global customers, while hitting high standards in development practices and ensuring GDPR compliance. Using Microsoft Azure also gives us additional deployment capability, to a common standard, across multiple data centres in different countries. Automotive companies are already using cloud technology across their businesses, for supply-chain, operations, manufacturing, and logistics, but they have yet to exploit it to connect vehicles and consumers. This may leave a gap in the market for the likes of Amazon, Android etc. who are creating the common platforms that are being embedded in the vehicles and link back to common cloud systems that everyone knows. We already have Alexa in retro-fit hardware, soon it may be built-in. While many OEMs are happy to wait for generic platforms, players like Volkswagen are walking up to the potential. It has invested 3.5 billion Euros in a new connected computing platform, vw.OS, which it aims to introduce in all its electric vehicles by 2020. This will connect embedded applications to handle vehicle functionality and provide consumers with connection to exciting new lifestyle services. It will be interesting to see how this battle for in-car cloud ownership plays out.
Interview from Smart Automotive Magazine ISSN2454-8561