Connectivity Hurdles Driverless Tech Still to Clear

Connected and autonomous vehicles are no doubt part of the future.

However, as with electric vehicles, there are hurdles that they are going to have to resolve rather than jump over to make them more efficient and reliable.  Peter Vermaat, head of connected vehicles and ITS at the Transport and Research Laboratory (TRL), nevertheless believes that the hurdles don’t emanate so much from connected, automated and autonomous vehicles.

In his view it’s more about a lack of services and the Catch 22 around connected services because nobody is providing connectivity, which he says is because nobody is providing the appropriate connectivity for them to operate on.

Challenged by standards

“We’ve had projects funded by the European Union (EU) looking at the viability of providing connected vehicles”, he explains before commenting: “There are other challenges around standards, and what kinds of connectivity to use. Nearly all vehicle-to-infrastructure (V2I) based services can be provided by a cellular connection, including 3G unless for streaming services. The problem is comprehensive of coverage differs for each provider.”

He adds that vehicle-to-vehicle (V2V) services are safety-based – such as automatic emergency braking (AEB), and non-time critical warnings of hazards up ahead that need to be clarified can be provided by V2I cellular services: “The problem is around is how you implement the V2V communications, and it was going to be done through dedicated short-range communications (DSRC), which is point-to-point data connection, but it’s like a bunch of people in a room and talking, speaking the same language. If nobody talks the same language or at the same time, then you have a problem. So, they speak a common protocol IEEE802.11P. Until about 2 years ago this was pretty much agreed that this was the way to go in Europe and in the US.”

To make the standards hurdle much higher to negotiate, he reveals that a competitor has come to the market promoting, namely vehicle-to-everything (V2X). The problem with V2X is that it speaks a different language and uses a different protocol. “This is being promoted by the cellular industry called cellular-V2X (C-V2X), and so what used to be clear has been muddied, he explains.

Vermaat adds that matters are being made more complicated because countries such as China have decreed to use C-V2X, while the EU prefers DSRC and in the US they used to support DSRC. However, he thinks the US seems to now be “going down the road of allowing the market to decide”. This illustrates that with standardization comes much politics. Countries as well as companies have different preferences over which standard to adopt. While standards are about finding common ground, not everyone wish to accept a particular standard preferred or offered by another party.

He, therefore, suggests that there will be a period of uncertainty, meaning that the required services won’t come forward. Yet, one standard will eventually win over the other to enable the two standards to talk to each other. He adds that it’s important to note that some safety services only “provide benefits when there is a high level of penetration of services (i.e. of about 50%).  Network coverage for V2I is not comprehensive too”.

5G potential

Kevin Roberts, senior mobility analyst at Ernst & Young considers the first challenge to set the standard for V2X and how vehicles will communicate with each other. He also talks about the potential for 5G mobile communications: “Much of the conversation currently focuses on 5G, however the 5G standard has yet to be formally finalized and while 5G has a lot of upsides, potential issues remain.”

He then poses some questions that will become increasingly critical for Level 4 and Level 5 of connected and autonomous driving, because connectivity is much are essential for everything to function: “For example, have you ever you’re your cell signal for a period? While inconvenient, it doesn’t post any major issues for the user, however if connectivity is lost to an autonomous vehicle what would happen? This becomes especially important if a Level 4 or Level 5 autonomous vehicles is utilizing Cloud or Edge computing. Secondly, while most of the focus is currently on 5G are there alternative standards that we should be considering like IEEE 802.11p. Thirdly, how are users going to pay for their data connectivity? Currently users pay on a per data used amount but how would you calculate the cost of data used between V2P or V2I?”

Connectivity and reliability

Sandeep Sovani, director, global automotive industry at ANSYS, highlights that standards are but one of the key problems. From his perspective, connectivity is the biggest hurdle: “For connected vehicles, one of the big challenges is to ensure that reliable of the connection. Unless like cellphone connections that drop, a situation that is a minor annoyance, when this happens with connected and autonomous vehicle it can jeopardize safety.”

He adds: “The vehicles will be using V2X communication, V2I, and V2V would be used when a vehicle is approaching a blind corner to detect vehicles that are around it. In such cases the vehicle will make split second decisions and, if that information drops due to a poor connection, then that can be a problem. For V2V connectivity has to be low latency, and very fast.”

For there to be any reliable connectivity, it’s essential to provide all the necessary infrastructure. From a 5G perspective, this means more cellphone towers than 4G used to make it all happen. Sovani concludes that speed to market is also a hurdle, or a challenge if you like: “If a company gets first to 5G, it will rule the 5G market for decades. It will, therefore, have a disproportionate share of the market. So, there is pressure to have the best time to market and the most reliable solutions to keep ahead of competitors.”

Environmental hurdles

He also believes that the environmental hurdles are not too different from those of existing cellular networks: “Research and current standards of radiation and electro-magnetic emissions suggest they are going to be enormous. However, what impact that has on the environment is open to question and, for example, the impact on honeybees is not proven.” Nevertheless, he warns that there is a need to consider solar storms in space because they can affect communications on Earth, which he finds is something that most people don’t even think about. So, it’s important to bear in mind all of the possible hurdles that could affect connected, automated and autonomous vehicle safety and the environment.

As part of the research process, the UK government is promoting test beds. Vermaat explains: “Zenzic are coordinating the building of connected and autonomous vehicles in the UK. For example, there is one at the Science Park in the Midlands. There will be test centers for off-road testing, for motorways and rural roads, and they will work to a common standard to allow manufacturers to test their systems. TRL has developed a safety standard for connected and autonomous vehicles, and that will be the same standard used for all test beds. This puts us at the forefront for testing.

“There is a lot of overlap between connected and autonomous vehicles. It is important to say that a connected vehicle isn’t necessarily autonomous, and an autonomous vehicle doesn’t necessarily have to be connected. However, autonomous vehicles benefit from being connected. Guaranteeing connectivity is not possible. The likes of Tesla will say that their vehicles are designed to not need to be connected. They will still be viable.”

Tips today for tomorrow

So, what are they their tips for overcoming tomorrow’s hurdles today? Roberts says the ecosystem needs to develop a collaborative mindset as many industries will need to come to market to make connected, autonomous, and automated vehicles happen. There is also a need to consider where you are in the value chain, and where you fit within it by also putting the consumer at the center of your thinking and developments. He also feels it’s important to regulate without hampering innovation and that there is an imperative to integrate mobility into broader urban policies.

Vermaat adds that because of the uncertainties around standards and the reliability of connectivity, there is a perquisite to be flexible: “Don’t tie yourself to one solution, technology or view of the future.”  This is because nobody yet knows what autonomous vehicles are going to look like 20 years’ time. He also warns that there should be some cynicism about the often-mentioned timescales for achieve Level 5 autonomy as: “We don’t know what mobility-as-a-service looks like, how it will work and even if it’s a good thing. In most cases it is.” With regards to Brexit, no matter how it happens, he says the UK is going to have to do it the EU’s way.

Meanwhile, Sovani recommends that the reliability of communications issue can be tested using simulation. This can enable automakers and other parties with a vested interest in the vehicles to test everything from the antennas used to enable vehicles to remain coupled together, environmental factors that may impact on communication reliability, different types of environment (city between skyscrapers, or rural), and simulation can also be used to test the reliability of chips or sensors. In effect he argues that simulations can reduce the time-to-time market by testing a wide variety of variables to make sure the systems are reliable than one would otherwise be able to do. In other words, simulations can assist carmakers to overcome today’s trials to overcome tomorrow’s hurdles.


Leave a comment

Your email address will not be published. Required fields are marked *