The move towards autonomous cars will herald a sea change that will reach far beyond the automotive sector alone.
What could be more exciting than a sector in which major car manufacturers such as Audi, Daimler, Toyota, BMW, Nissan, Volvo rub shoulders with new electric vehicle manufacturers such as Tesla and are also vying with established tech giants such as Google, Baidu, Apple, Samsung, Tencent and competing with new tech such as ride hailing companies such as Didi and Uber?
What could be more exciting? Well the impact of that other great disruptor – blockchain – in autonomous cars that is what is even more exciting! Blockchain technology, like autonomous cars, has the ability to disrupt cosy ways in which business has been conducted for eons. However, in the ecosystem of autonomous vehicles blockchain may not be a disruptive technology but rather an enabling technology. Cybersecurity is one of the major challenges that autonomous vehicles face. Indeed cyber-security may be the Achilles heel of autonomous cars – if this chink in the armor cannot be addressed then it may well be “bye bye global roll out of driverless cars”.
However, due to its very essence blockchain may play a critical role in solving or mitigating cybersecurity risk (i.e. an “Achilles Boot”). In addition to solving this critical issue blockchain will intersect with autonomous vehicles in areas such as data sharing, payments, insurance, ride sharing etc.
Autonomous cars, also referred to as being driverless, self-driving, robotic cars are able to navigate without human input. In order to do so such autonomous cars need to utilize integrated technologies such as radars, LiDAR, GPS, odometry, computer vision and hardware and software. These technologies allow the car to accurately assess and navigate its environment. Cars will no longer be self-contained but rather seamlessly connected to roadside infrastructure, with each other and traffic management systems. In this way autonomous vehicles will fall within being part of Internet of Things (IoT).
Elimination of human input in the cars and delegating control tasks to hardware and software and creating an interconnected web of cars and traffic management centers revolutionizes the transport and automotive sectors (and much more) but also creates a number of vulnerabilities to car systems that can be compromised by the malevolent. These vulnerabilities are not theoretical but real threats. Hackers could gain control over core functionalities of the vehicle including the braking and accelerating systems, control of steering or just shut down certain components of the car or shutting down the car completely.
Conventional digital infrastructure of smart and autonomous cars will result in vulnerabilities being spread across the centralized architecture of the autonomous system as well as allowing privacy issues and safety threats to spread. In addition a conventional digital infrastructure will have limitations in the scalability of the technology.
There are three main vulnerabilities connected with conventional security architecture:
Centralization: Centralized architecture of the interconnected web for autonomous cars is not well suited for this industry. Most importantly the technology has a single point of failure which can then cause suspension of service and result in unresponsive servers (such as DOS attacks). These would in turn lead to major security breaches in the central database which would allow manipulation of the whole data and disruption of the entire network. The current architecture relies on centralised brokered communication models where all vehicles are identified, authenticated, authorised, and connected through central cloud servers. This infrastructure is unlikely to be able to scale as an ever increasing number of vehicles are connected to one another and the cloud.
Privacy: The current secure communication systems gives rise to privacy concerns as systems collect sensitive private information from all-encompassing database without the permission and knowledge of the data subject. Concerns go beyond private data being compromised in case of a security breach but such data is also inefficiently summarized and sold off to third party service providers.
Security breaches: As noted before the main threat is for a hacker to compromise the security of the system and exercise control which will threaten the safety of drivers and/or passengers.
Security Solutions Provided Through Blockchain
Fortunately, many of the vulnerabilities in the conventional safety and security architecture can be solved through deployment of blockchain technology. For most people blockchain is known as being the underlying technology for cryptocurrencies such as Bitcoin. However, it is much more. Blockchain is a distributed peer to peer network relying upon distributed ledger technology and smart contracts in which all communications between nodes is encrypted with public keys and broadcast to the network. Periodically one of the nodes is chosen to create a block and combines all the pending transactions into a block which is added to the blockchain and broadcast to the whole network. Once a block is added to the blockchain it becomes immutable and cannot be changed by hackers. These properties of blockchain, namely being decentralized, cryptographically secure, immutable and traceable may be a solution to some of the challenges associated with autonomous cars.
Decentralized: Blockchain technology is based on a decentralized architecture, which allows a massive number of nodes (computers) to participate in reaching a consensus on the state of the database. The decentralized architecture eliminates the threat of a single point of failure. This is because even if the majority of nodes are not operative there will always be a few nodes which are able to keep the database running.
Immutable: Once a transaction and data has been verified through consensus and added to the distributed database it is impossible for a hacker to amend or substitute the blocks in a blockchain.
Encrypted: Although blockchains are transparent in that stored data is readily accessible to the public to monitor the combination of data storage and simple encryption allows for data to be only accessible by the owner of such data or by parties which are provided access by the owner.
Secure: A major security problem of conventional technologies is the threat of hackers installing a modified, malicious software update which easily compromises the affected systems. Blockchain through usage of public key and private key functionality of cryptography, can require every software update to be signed by the exclusive private key which only the driver and software developer have access to.
No Trust Needed: A by-product of the consensus algorithm of blockchain, transparency and its immutability properties means there is no need for a central trusted authority in the system. Counterparties connected by way of a blockchain network can interact with one another in a trustless manner.
Reduced latency: Blockchains have many local nodes involved in their architecture which can allow for faster data collection and data transfer.
Figure 1- A Summary of Blockchain advantages to conventional methods employed for studied applications (source Dorri, Steger, Kanhere, Jurdak (2017))
Blockchain as Part of the Autonomous Vehicle Ecosystem
The perfect storm of the parallel development of autonomous vehicles, distributed ledger technologies and IoT will radically change the face of transport.
Likely ways in which the blockchain (i.e. distributed ledger technologies) will be used in enhance autonomous vehicles are as follows:
1. Data Sharing
Data sharing is one significant example of how distributed ledger technology can enhance safety in respect of autonomous vehicles.
Autonomous vehicles rely on a series of sensors, including GPS to navigate, LiDAR to avoid collisions as well as an array of optical cameras. In order to be driverless the vehicles have to be basically enveloped in data and will also need to feedback data in order to strengthen the network of autonomous vehicles. Blockchain powered data pooling would improve safety, efficiency and convenience. Examples would include flagging intersections or blind corners where there is a high incidence of hard braking. This flagging would serve as a warning to autonomous vehicles to take extra caution in such area. Blockchain technology can additionally be utilized to transact real-time data packets and incentivize car owners to share data and thereby strengthen the overall vehicle network. Map platforms like Google Maps will likely purchase traffic information on expected delays around a new construction site and local council authorities may purchase information on the location of potholes that have been detected.
Blockchain coupled with cryptographic encryption enables relevant portions of data to be only accessible by authenticated parties, such as pothole locations for council authorities. The blockchain would guarantee the vehicle as a legitimate source of data while keeping details such as registration number, passenger identity or vehicle speed secret. Already today there are a number of companies using blockchain to enable data sharing. For example Toyota Research Institute (TRI) uses blockchain to connect the varied onboard sensors used by vehicles while allowing consumers to own “their” data. TRI has partnered with BigchainDB to build a data exchange to allow for driver data sharing. Streamr is creating a decentralized data marketplace that allows users and machines to handle data in a completely P2P environment.
2. Car Sharing
For many people the family car is the biggest investment after the family home. In addition it is a major expense that is only used for a few hours in the day – and to add insult to injury in large cities most of the time spent is in gridlock rather than actually getting anywhere – for example in London drivers spend more than 100 hours a year in gridlock. In addition a major headache is finding a park once you get to your destination. Various report estimate that cars are in motion only 4% of the time and remain idle most of the time. Despite this low utilization rate car ownership remain high. In 2015, there were 82 registered person vehicles for every 100 people in the US. The sheer volume of stationery cars has serious implications for urban planning as large areas of urban space are occupied by carparks.
Car sharing is a recent development which addresses the above inefficiencies in respect of mass private car ownership in cities.
Autonomous vehicles will catapult car sharing from hipster eccentricity to mainstream practice. The convergence of the sharing economy coupled with distributed ledger technologies will enable cars to remain constantly active – offering transport to paying passengers. These constantly on the move autonomous cars will reduce the need for carparks as less vehicles will be standing around idle, reduce the need for personal car ownership and will provide supplementary income for car owners.
Distributed ledger technologies will facilitate mainstream adoption of car sharing by scheduling and matching rides without the need for a middleman. Distributed ledger technologies can allow information on vehicle availability to be made publicly accessible so that users and car owners can match journeys easily. Scheduling is particularly important for car sharing platforms offering one-way trips such as ZipCar and DriveNow by BMW as they often involve return trips to urban areas.
The decentralized nature of blockchain would also enable drivers and passengers to make P2P bids without centralized mechanisms that sets prices or takes a fee such as Uber. Indeed blockchain may disrupt the disruptors as consumers and service providers can meet without the need for a centralized bureaucracy. An added benefit of using a blockchain powered ride sharing system would be to enable drivers to keep the 20-25% fees that are typically charged by platforms such as Uber and Lyft. TRI is currently collaborating with MIT, Oaken Innovations and Commuterz to create a unified platform for carpooling and car-sharing. Oaken is also working on building a system for short-term vehicle leasing.
Another way distributed ledger technology could support autonomous driving is to enable vehicle-related payments without human interaction. No-one enjoys refueling, recharging, maintenance, parking – so why not let the vehicle do these things on its own – and also pay for the services at the same time.
Beyond facilitating daily operational micro-transactions, distributed ledger technologies may also offer a solution to the difficulties in storing solar-generated electricity. Rather than being distributed via a centralized grid the blockchain technology would allow excess electricity generated by nearby solar panels to be sold on a P2P basis to passing by automated vehicles. This would greatly reduce the cost associated with electricity being lost in transmission around a grid. This is not a pipe dream as ElaadNL, an innovation center in the Netherlands, is currently testing just such a system. In addition BlockCharge is a blockchain based platform which provides a fully automatized, efficient and fast vehicle charging infrastructure. Share & Charge has developed a decentralized application to enable users in Germany to share their private car charging stations and set their own rates for usage. Oaken has developed a proof of concept to enable Tesla cars to pay for road tolls over the ethereum network.
Accordingly many of these applications are already being utilized today.
Car insurance is not generally thought of as being innovative. However, autonomous vehicles coupled with blockchain based advanced data sharing functionality and payments will allow for new forms of car insurance models not previously possible.
Currently car insurance models are based on yearly premiums even though this has no relationship as to how much the car is in operation or indeed at risk.
Blockchain will make it possible to move to a usage-based insurance (UBI) or ‘pay as you drive’ (PAYD) system. Before fully autonomous driving becomes a reality data collected from advanced sensors can be compiled and put into a blockchain which would reveal driving behavior and habits of individuals. This would in turn allow insurers to give discounts or better premiums for good driving. While blockchain is not the only solution for storing this information, the immutable and auditable nature of a blockchain ledger provides increased transparency and significantly reduces possibility of fraud. This in turn will allow insurers to lower premiums.
When autonomous vehicles achieve full autonomy (Level 5) then insurance may be calculated on a pro-rata basis along with a mileage fee per ride. Blockchain and smart contracts could be further utilized for insurance claim processing, by automating claims submission and payouts and automatically deploying service technicians. Gem and Toyota are currently working on usage-based insurance products tied to the telematics that are being produced by a user’s vehicle. These companies provide a ledger for distributed inputs from a number of different sources which can then be used to automate much of the insurance claim process.
5. Vehicle History
Distributed ledger technology is also being applied to create immutable, verified ledgers of a vehicle’s history. Although this is a useful application for traditional cars, it is particularly useful as we move towards autonomous vehicles.
Shared car use will rise relative to personal car ownership as a natural result of autonomous vehicles. As more people move to car sharing there will be increasing numbers of consumers interested to know details about the vehicle’s history i.e. kilometres driven, any accidents etc. Innogy, a subsidiary of German energy company RWE, has announced plans for a blockchain platform that creates a digital ‘twin’ of the car which will record details provided by the manufacturer, repair shops and the car itself and which will be stored on the IOTA tangle. Through blockchain data verification, passengers and owners will be confident that odometer readings have not been tampered with and that any accidents or defects have been dutifully recorded.
There are a range of challenges in establishing an IoT ecosystem which uses blockchain to connect autonomous vehicles, public utilities such as toll gates and charging stations and third-party data purchasers. Each additional function strengthens the network and makes vehicles more autonomous. They also allow the vehicle to become more autonomous beyond mere self-driving capabilities. However, this requires coordination from a range of parties, including stakeholders such as regulators, vehicle manufacturers and high tech companies. Again an issue will be that technological development will very likely speed ahead of the legal framework.
In addition to the above legal issues the use of blockchain technology in autonomous cars will raise novel legal questions which are yet to be answered.
One major legal issue will be assigning responsibility for glitches or other technical errors. As all distributed ledger technologies are decentralized there will be no single centralized body that maintains the network.
In micro-payment situations the circumstance could arise where an autonomous vehicle charges its battery but is over- or under-charged for electricity. In this case it will be unclear whether the responsible party is the autonomous vehicle manufacturer or the charging station. Indeed, in situations where glitches occur due to errors in the distributed ledger technology itself, as opposed to participating people or devices, it is unclear whether the original developer itself may be liable.
Another legal minefield for blockchain will be in respect of data privacy. Regulations may require drivers and/or passengers be given the option to customize which data is transmitted to the data marketplace and which is masked, particularly in case of shared autonomous vehicles which have large numbers of passengers who will have different privacy preferences.
There is no doubt that blockchain and distributed ledger technologies have much to offer for the development of autonomous vehicles. Blockchain will enhance user experience, reduce inefficiencies in electricity transmission and reduce the high rate of private car ownership through promoting car sharing.
Blockchain has the potential to fundamentally change:
- the way autonomous vehicle data is collected, stored and transacted, by incentivizing it, securing privacy and immutability,
- the way car sharing solutions are operating, by offering solutions for intermediary-free ride hailing, car sharing, carpooling, leasing services,
- the way payments and micro-transactions systems are designed, through usage of smart contracts and trustless transactions,
- will affect the way car insurance products are designed, allowing innovative insurance products, bringing, efficiency, transparency and speed into the car insurance process,
- the methods that information and history of a car is stored, enabling a transparent database that could give an elaborate account on any given car.
A wide range of blockchain and non-blockchain powered platforms, think tanks and research institutes have proposed platforms to address a number of needs in this space (a sample list is provided below). However, despite this rapid development, it must be remembered that there remain potential flaws with the underlying technology as well as unanswered legal issues regarding privacy and liability for technical errors. These issues will need to be addressed before the technology is implemented in a mass market manner.
A sample of companies and start-ups operating in the industry (including blockchain powered solutions)