Telecoms.com periodically invites expert third parties to share their views on the industry’s most pressing issues. In this piece James Cater, EMEA VP for Spirent Communications, takes a look at the implications of 5G for the transport and logistics sectors.
5G and connected technologies are going to bring about a revolution in transportation.
While automated transportation has been theoretically possible for a long time – mobile technology has not been capable of living up to the promise of self-driving vehicles. However, the arrival of 5G is dragging these possibilities into the real world. As such, manufacturers will need to consider the digital realm like they’ve never done before.
Soon our goods could be transported across the country by automated trucks, drones could arrive at our doors with our packages and we could be chauffeured around the cities in the very near future in driverless, fully automated vehicles. However, that future isn’t set in stone. None of these technologies – nor the 5G infrastructure which underpins them are forgone conclusions.
While they represent transformative new innovations they also present new risks and new requirements. Automated vehicles – and the connected components which make them possible – will need reliable coverage and connections to realise their capabilities, as well as an incredibly high level of data accuracy and safety. Without that, automated vehicle technology won’t be viable.
There are many issues to deal with in the design and development of automated vehicles and many potential potholes on the road ahead.
Positioning, Navigation and Timing (PNT)
Accurate vehicle positioning is already a reasonably common piece of technology found in ride-sharing apps. However, automated driving is going to need much greater precision. These vehicles will need decimetre level accuracy on motorways and centimetre level accuracy on residential streets.
To do that, they’ll rely on a variety of different sensors, all of which operate reliably to ensure the requisite accuracy. GPS and GNSS sensors will determine the absolute location of the vehicle on the surface of the earth. Lidar, radar, and sonar will detect objects in the environment, read signs and inform various safety-critical drive decisions. Accelerometers and gyroscopes measure speed movement and orientation. These are just a few of the complex array of sensors that automated driving will require to remain safe.
The data from those sensors will then be fed into the Advanced Driver-Assistance System (ADAS) a software application that will then synthesise that data to build a picture of the vehicle’s environment and response in real-time.
Vehicle to Everything (V2X) Communication
Automated Vehicles will constantly talk to their drivers, each other and everything around them. One of the things that makes the coming transportation revolution possible is that 5G can offer the performance and coverage to make that critical level of communication possible.
Automated vehicles will need V2X to operate safely on the road, to understand where they are in relation to other objects and vehicles and ultimately to keep their drivers, passengers and contents safe.
With that in mind, V2X communication must work in any and all conditions that the automated vehicle might face. Furthermore, it must work independently of the Radio Frequency Technology or Protocol stack that’s used within the vehicle.
Time-Sensitive Networks (TSN)
Automated transportation – especially private cars – will be replete with applications that need to send, receive and access data over an ethernet network. That includes systems such as infotainment and automated driver assistance to on-board diagnostics. In the meantime, they’ll have to perform potentially millions of application transactions per second.
Many of those systems and applications are time sensitive and critical to the broader safety of the vehicles and the driver. As a result, vehicle manufacturers have to ensure that those in-vehicle data services are both reliable and secure.
Cybersecurity
In 2015, security researchers showed the world that they could hack a Jeep Cherokee. In the aftermath, Chrysler was forced to recall 1.4 million vehicles.
Automated vehicles are made possible by the mere fact that they are connected to the internet. This also throws them into a new risk landscape populated with digital threats. For example, GNSS signals can be remotely spoofed and cybercriminals can already hack vehicles through insecure infotainment systems. A hack on a vehicle is a troubling possibility and manufacturers will have to get to grips with a new category of risks.
Bold possibilities and important considerations
The 5G powered automotive revolution brings new possibilities as well as new considerations. Its promises are bold and the complexities it brings are deep. The introduction of software and radar technology into largely “offline” automotive technology is going to require developers, network operators and vehicle manufacturers to think in new ways.
Those potential potholes and complications need to be worked out thoroughly before these vehicles get to market. That’s why testing and assurance has to be a central part of the development lifecycle.
Testing Automated Vehicles
Before any of the goals of this 5G empowered transportation revolution are realised, stakeholders will need to be sure that they can fulfil them. Those vehicles have to be reliable and safe before rubber ever hits the tarmac.
Testing automated vehicles will involve millions of miles and the wide variation of conditions that any one model of vehicle will have to drive in a consumer’s hands.
That goes for every individual part of the vehicle too. For example, each individual sensor needs to be tested to make sure that it’s providing accurate data and that it’s combining with the algorithm to create an accurate picture of the car’s environment. GNSS can lose signal in tunnels, be compromised by tall buildings or even be cyberattacked. All of those possibilities will have to be thoroughly accommodated within testing.
The problem with testing on the road
The most obvious option might be to merely test these in-development vehicles on real world roads. However, real world testing cannot hope to provide the amount of data required to feed into those automated systems.
A test vehicle would need hundreds of millions of miles in all kinds of environments, locations, weathers and situations to gather enough data for real-world use. Furthermore, this kind of testing cannot simulate the edge-cases or widely varied conditions in which safety and reliability are so important.
Many manufacturers say that they’re going to have automated vehicles on the road by the middle of this decade. Real-world testing won’t get them there in time.
Testing in the lab
Those kinds of conditions and requirements can only be simulated in a lab. Manufacturers are now using labs to aid their automated vehicles clock the millions of test hours they’ll need to validate their technology’s reliability.
Lab testing can imitate any possible condition that a given vehicle might face in the world and provides a method which is fast, scalable, reliable and repeatable.
Transportation is transforming and in a few short years, we may see driverless cars, trucks, drones and more on our highways and in our airspace. Their capabilities are underpinned by the connectivity that 5G permits and so the availability and integrity of connected components is absolutely critical. That dependency means that automotive manufacturers will require disciplines and standards that manufacturers have never had to deal with before.
The road could be long and arduous, but to realise these possibilities they’ll have to make absolutely sure that these vehicles are reliable and safe. The viability of automated driving is not a foregone conclusion but thorough testing is a step towards it.