Peter Stoker, Chief Engineer – Connected and Autonomous Vehicle at Millbrook, explains the role that 5G is playing in the development of connected and autonomous vehicles
Collaborate to innovate
The 5G testbed for transport at Millbrook Proving Ground, launched last year as the AutoAir project, is a private, fully operational high-speed mobile data network. It was installed to support the development, testing and validation of connected and self-driving vehicles.
It is through my involvement with this project that I have come to experience the ground-breaking work being done in the field of connected and autonomous vehicle (CAV) development.
As the first network of its kind in the UK, AutoAir really is leading the charge when it comes to innovation. Not only is it supporting developers of CAVs and associated technologies, it is also helping to position the UK automotive industry as a leader in global CAV and driverless vehicle technology development.
Before we dive too far into the use cases being explored and the impact that AutoAir is already having on future technology and transport infrastructure, it is important to first understand the origins of the testbed.
In 2017, the UK government Department of Digital, Culture, Media and Sport called for the establishment of 5G vertical sector testbeds and trials. The AutoAir consortium, led by Airspan, which brings together leading lights from the mobile communications and transport sectors, was formed in response to call to action.
The testbed is the only accelerated development programme for 5G technology based on small cells that operate on a neutral host. This makes it a truly unique set up. It allows multiple public and private mobile network operators (MNOs) to simultaneously use the same infrastructure using network slicing, which can radically improve the economics for 5G networks.
As part of the project, the consortium set up 60GHz mmWave mesh radio between small cell sites to connect them to the core network (‘backhaul’). This has enabled the consortium to compare this with the costs of deploying fibre.
The testbed itself consists of 89 radios, covering 2.3, 3.5, 3.7GHz 4G and 5G spectrum, 60GHz mmWave mesh and 70GHz high-speed vehicle-to-infrastructure links. 59 masts were fitted around Millbrook, linked by 30km of power lines and fibre cabling.
What have we learned so far?
The AutoAir testbed has already yielded significant insight. For instance, it’s provided clarity as to how MNOs, vehicle manufacturers, governments and transport operators could harness neutrally hosted 5G and mmWave spectrum networks in the future for a more cost-effective and connected mobility.
AutoAir’s innovative proposition is a wholesale access neutral host hyper-dense small cell deployment model for transport corridors. It provides a single, shared infrastructure set across multiple MNOs. This makes mobile services on transport corridors more attractive for mobile operators and end users, unlocking a multitude of possibilities.
For example, in the UK, all four existing MNOs would be able to share the same physical network. In addition, other organisations, such as emergency services, road maintenance firms and vehicle manufacturers would be able to run their own private networks on the same shared infrastructure at a fraction of the cost of deploying their own physical networks.
What are the use cases explored so far?
It should be evident that the fledgling stages of the AutoAir testbed were more concerned with transport infrastructure. The reality, though, is that the AutoAir testbed has only really begun to scratch the surface of how 5G technology might be harnessed more widely in the automotive sector.
That is why it is exciting, and hugely important, that the AutoAir testbed is now being operated on a commercial basis. This gives CAV developers the ability to really push the network to its limits and make significant advances in their technology.
As a result, a variety of organisations, working on a myriad of uses-cases, are already exploring the capabilities of the 5G network. One particularly interesting, and potentially lifesaving, trial that was successfully run courtesy of AutoAir was the ‘Smart Ambulance’ trial with the East of England NHS.
This pioneering project involved equipping a standard ambulance with state-of-the-art devices and connectivity to create a Smart Ambulance that simulated 5G connectivity. The ambulance was transformed into a unique remote consultation room, able to relay a live video stream to a remote team – potentially saving the time needed to save a life.
And that’s just one example of how the super-fast data transfer afforded by 5G might shape our futures on the road.
Indeed, the 5G testbed at Millbrook is also enabling CAV developers to expedite the testing and development of new infotainment and multimedia technologies. As was demonstrated with the Smart Ambulance trial, 5G facilitates vehicle-to-vehicle (or vehicle-to-remote location) communication in real-time.
But that’s just the tip of the iceberg. This new level of connectivity enables over-the-air software updates in real-time, as well as delay-free video and music streaming, real-time map downloads and more.
The spectre of cyber attack
A word of caution. This new age of connectivity and autonomy is not without its pitfalls. One of the biggest areas of concern in relation to connected and autonomous cars has always been cyber security.
With enormous amounts of data being transferred in real-time from vehicle to vehicle, vehicle to infrastructure and beyond, there is little wonder there are concerns over hacking and privacy.
It is therefore essential that providers must also be able to conduct physical, on-road tests to prove system robustness in a safe and controlled environment. This is all the more important when you consider that, for vehicle manufacturers, future type approval will require evidence of cyber security assessment processes.
Valuable insight into tackling the threat of cyber-attack has already been gained – thanks in no small part to Zenzic and its associated organisations – but there is still much to be done. Indeed, security considerations of the CAV ecosystem will require ongoing research and assessment as the technology and threat landscape change.
And, let’s face it, they will keep changing.
How exactly could a security breach manifest itself? The simple answer is that there are several security threats that can have an impact.
This can include disruption of connectivity to prevent operation of CAV services (denial of service), data privacy issues (including data exfiltration and tracking), and the planting of malicious malware within equipment, both at source and by threat actors.
Evidently, the threats posed by a 5G cybersecurity breach are potentially far more serious than those of previous generations of the technology, such as 3G and 4G.
The harsh reality is that, just as 5G is a step up in capabilities for its users, with enhanced broadband speeds, latency and connectivity advantages, those abilities also aid threat actors. They can extract data sets from systems quicker, increase the capabilities of botnets, and target more devices as 5G IoT devices become widespread.
Those are the threats, but what about the potential ramifications of a cyber-attack for the average road user?
As we’ve learned, with vehicle systems becoming more sophisticated, the possibility of undetected vulnerabilities existing within those systems increases. Furthermore, increasing commonalities of vehicle components and use of commodity operating systems means that manufacturers and researchers must guard against the possibility of attacks causing mass disruption of vehicles.
Meeting the challenge
The industry is becoming aware that cybersecurity is not an end goal but an ongoing arms race, and cyber-resilience is the key.
It is important that we maintain an up-to-date understanding of the current threat landscape for the whole system. It is also essential that we keep revising security of the mobility entities through testing and analysis, especially when new breaches are reported.
The lifetime of any security certification may need to be continuous, or at least only be valid for short timeframes – say, six months.
