What Equates to a Smarter Automotive Future?

In Britain there is much derision for so-called ‘smart motorways’, with drivers often considering them to be both dumb and dangerous.

Most motorists will have a large list of being caught in lengthy traffic jams because of these connected highways that lack the simple expedient of a hard-shoulder for vehicle breakdowns resulting in these first responders having to fight their way through miles of static traffic. So, there are many questioning just how smart is the future of connected and autonomous vehicles actually going to be, and considering the public’s criticism of smart motorways, how the industry is going to overcome any cynicism about the technology?

Robert Camm, senior consultant of mobility at Frost & Sullivan, responds by saying that the intelligence of the latest connected and autonomous vehicle is at a good level. “As well as the capability that the driver sees, there is also a lot of intelligence running in the background that may not be seen,” he explains before giving some examples of seen connected and autonomous technology.

They include such things as: connectivity, live traffic feed with route optimisation, swarm data with vehicles sharing information on incidents, road conditions such as icy highways and vehicle-to-infrastructure (V2I) in selected markets to share information such as traffic light data to the driver.

Autonomous driving technology

An increasing number of vehicles now have Level 2 and Level 3 autonomy. This supports hands off driving on the highway using high-definition mapping in the background to provide a real-time 3D map of the environment around the vehicle.

Camm said: “This can be used as a reference for the autonomous driving algorithms. In addition to the on-board sensor data, in the event there is a change in the environment, the on-vehicle sensor data can be used to update the map and this can reflect to other vehicles.”

He says this is already happening for vehicles, especially those such as Ford BlueCruise, or Mercedes DrivePilot. They require a mapped road as a prerequisite for their systems to work. In addition to this, he explains, there is the need for a data exchange between the and cloud in real time. He says this requires advanced infrastructure and data processing algorithms. Key to this is the ability to process data and provide feedback to the vehicle.

‘Big-Loop’ data

Manufacturers such as Volkswagen are terming this ‘Big-Loop’ data. He adds that in this context, the growing intelligence that can be achieved today with connected vehicles can be considered smart.  Camm, nevertheless, agrees that there may be some issues with the infrastructure away from the vehicle, or with the automakers’ ecosystems, which may be causing a disconnection and producing less intelligence. Yet sometimes, it’s just about the ‘smart’ bit is about how the technology is applied.

Miro Adzan, general manager of ADAS at Texas Instruments adds that today’s connected vehicles already use information from a variety of sources, which are internally and externally sourced. External data sources can be used to create a better traffic experience and, so, this includes information from other vehicles to identify congestive areas, road blocks and other highway events by analyzing the speed of other vehicles.

He said: “Autonomous cars already achieve level 3 on highways. Some OEMs have introduced Level 3 solutions on highways up to 60kph (37mph) and will introduce solutions in 2023 for up to 130kph (81mph). Additionally, valet parking is happening worldwide where the car can find its own place in the parking garage with onboard sensors and with sensors, hence connected, in the garage.”

As for the lessons that can be learned from smart motorways, Camm comments: “Smart motorways refers to motorways that have a virtual hard shoulder to improve traffic flow. They also have variable speed limits deployed more regularly to slow down traffic if there is congestion ahead, or if there is danger or obstruction in the road. The benefit of them is not yet entirely clear and with hard shoulder accidents the effectiveness is questioned.”

Smart motorway dilemmas

One of the key issues, he says, is the time it takes to develop smart motorways. Almost as soon as they are completed, they become outdated. This means the technology has to be replaced by solutions that are more intelligent and connected to the vehicle in the future. The consequence of smart motorway technology becoming outmoded by the time a motorway is completed is that they have often been scrapped, owing to budget, and because to a lack of driver confidence in them.

He adds: “However, this time delay is what creates a challenge with V2I as a whole. Where infrastructure technology roll-out is much slower than on-vehicle technology. Which is why, for now, autonomous vehicle systems are not incorporating infrastructure sensors or data to operate.”

To this question about smart motorways, Adzan comments about how ‘smart’ applies to the vehicle. He says intelligence in a vehicle is not a value of its own. The features and functions that are implemented must add value to the driver rather than be a distraction. He says autonomous and smart functions need to be adjustable and dis-engageable: “So that different users can adjust what level of functionality to use as it permits individuals to use the functionality they are fond of and see value in.”

Smarter automotive reality

So, what equates to a smarter automotive future, what are the hurdles and how will they be overcome to make it a reality? Camm thinks the future will have enough intelligence on the vehicle to deliver smart benefits to deliver road safety improvements. As for infrastructure, the technology should be about communicating the most important information, such as speed limits of lane closures.

He suggests that other information such as road conditions, traffic congestion and so on, can be communicated V2V through crowdsourcing. He adds: “I have already seen this with navigation systems such as Waze but in future it will be built in and largely automated. For example, in the event of a lane closure it could be communicated vehicle to infrastructure, but also vehicle to vehicle. So that the driver, or the autonomous vehicle, knows to leave the lane well in advance.”

In the event of change in speed limit, the vehicle could respond by automatically changing speed. This can already be achieved with some cars. Other vehicles could also receive communications from that car, truck, or bus, to pre-warn other vehicles of the change in speed ahead.

Hurdles to overcome

Despite these benefits, Camm says there are still some hurdles to overcome. They include data standards, what information is needed and how is it communicated; communication standards such as Bluetooth, Wi-Fi, 5G – which will be used for vehicle-to-everything (V2X) communications; vehicle standards; and there is a need to ensure new vehicles are fitted with increased safety equipment.

He adds: “EU general safety regulation is a good example of how the bar is constantly raised on technology mandated for vehicles, and there is also the need to manage the transition between vehicles as there will be periods with autonomous vehicles running alongside ‘manual’ vehicles, which will reduce the overall safety benefit.”

Adzan says: “Intelligence in cars will come from smart solutions inside car and smart solutions outside of the car, as well as the combination of the two,” This is because intelligence in the car will be mostly about what data the sensors in the car, or vehicle, can obtain. There is also the question of how much intelligence the on-board processing will bring.

He explains: “This is especially true for camera and radar data that is captured. The resolution of cameras and radar sensors is increasing, as well as the amount of sensors outside and inside the car. One example of this is combining the driver-facing camera with other sensors on the wheel or on the brakes. These combined sensors can identify if the driver’s eyes are shutting when sleepy, or if the driver is “jerking” the wheel when agitated, angered or when driving too close to another car.”

Driving support

Information like this can be used to program the systems that support driving. “A driver who only steers when very close to another car and then abruptly might be alerted with a sound much earlier than when the driver is behind the wheel who is more attentive and steers earlier in the first place,” he suggests.

Like Camm, he finds that the data communication between the vehicles and the infrastructure needs to be stable in every driven area. He also stresses that it needs to be affordable and timely. In terms of affordability, Adzan says the sensors need to be ‘cost-optimized.’ So, a smarter automotive future is essentially about having the right technology in place in order to use and analyze data for the benefit of drivers, connected and autonomous vehicles and to gain the ability manage traffic and road safety much better than has been possible in the past. This must be done intelligently to avoid being dumbed down in the way that many members of the public perceive smart motorways.