Autonomous Long-Haul Trucking to Solve Driver Shortage

By Graham Jarvis

---

30th November 2021

At the end of September 2021 media reports in the UK suggested that there was a shortage of drivers to deliver diesel and fuel to service stations gas stations.

The fuel crisis that impacted the UK for almost a month and which, according to the Petrol Retailers Association, saw nearly 5,500 independent petrol outlets run out of fuel, with many more running partly dry. Karl Simons OBE, executive director for health, safety and wellbeing at FYLD, commented on the issue: ““It is inevitable that the current fears over fuel supply will recur in the future. However, with the digitization of systems to predict and control risk, we can expect more accurate forecasting of consumer demand, predicted shortages of essential products or services and the utilization of key personnel to ensure the most urgent task or highest risk is prioritized. Only then can these crises be averted.”

Joseph Salem, partner and head of transport for the Middle East and India at Arthur D. Little, adds: “The clear message is that it is critical to have alternative power sources, whether you are looking at electrification or the conventional powertrain. It’s important to accelerate the push from away from petrol and diesel by fast-tracking the policies to promote the adoption of zero emissions vehicles – including electric vehicles and hydrogen-powered vehicles in the longer run.”

Meanwhile, on social media people suggested this would give impetus to electric vehicles, while others suggested that the crisis highlighted the need to rely less on human drivers by pushing forward the benefits of connected and autonomous trucks.

Autonomous trucks

Ann Steffora Mutschler argues in her September 2021 article for Semiconductor Engineering: Long-Haul Trucking with Fewer Drivers, that the economics for recruiting CAV trucks is certainly compelling. However, she also suggested that there are some significant technological challenges to overcome that are unique to the trucking and logistics market. She also suggests that the industry is becoming very much dependent on “increasing levels of autonomy and electrification to reduce the cost of moving goods and to overcome persistent problems”.

Robert Day, director of automotive partnerships at Arm says the trend towards increasing automation and autonomy in vehicles is clear. “Areas like haulage and large truck fleets are interesting because along with applications like robo-taxis or autonomous shuttles, we are closer to mass deployment and nearer-term commercial reality than privately owned, fully autonomous vehicles, which are still some years in the future.

“This trend is driving the growth of flexible SoC platforms that deliver efficient, high-performance compute coupled with heterogeneous processing, computer vision, and sensor fusion. It’s critical that these platforms are built on robust and safe foundations, so our focus with these partners is on creating solutions that combine the right levels of functional safety with scalable, specialized compute that addresses the range of processing capabilities required for safe autonomous capabilities.”

Deployment at scale

Day finds that there are other clear signs from the market to consider. There is, for example, a strong desire to deploy autonomous vehicles at scale. The desire for increased autonomy is also creating a push for strong partnerships between the truck manufacturers, autonomous technology companies, and logistics companies, with the aim of deploying autonomous Class 8 trucks.

He adds: “Recent publicly announced examples include Volvo Autonomous Solutions working with Aurora to develop and commercialize L4 Class 8 trucks, a collaboration between FedEx, PACCAR and Aurora to bring autonomy, trucking and logistics together, and Navistar partnering with the TuSimple autonomous technology company, to commercialize self-driving Class 8 trucks at scale.”

Day also agrees that the economics around greater efficiency in terms of miles and hours driven per driver are compelling when much of the driving is done by an autonomous system. This is because companies will be able to make better use of their assets in terms of the vehicles themselves and their drivers. The key benefit is the gaining of extra potential revenue per truck and per driver.

He explains: “With more autonomy, the truck and driver will be able to drive further, which presents both opportunities and challenges. Most autonomous systems today are created with a particular operational design domain (ODD) which equates to the environment that it is designed to operate in, including speed, diverse environmental conditions such as in different kinds of weather or times of the day and also in different geographic region.”

Technological challenges

However, he warns that as the distance travelled increases, the autonomous system will have to broaden its ODD, which often leads to more technological challenges that need to be solved. “For example, a truck that is travelling west to east across America will have to drive in both day and night conditions and may encounter different weather conditions such as heavy rainstorms in the South-East,” he explains.

Subsequently, he says: “The vehicles have to learn more about the dynamics of a wet road and so they will also potentially have sensor blindness caused by heavy rain.” This means that the sensor modalities and the compute system will need to change “from a system that just operates in the sunshine and daytime” to one that can operate no matter the time of day or weather conditions. So, as long-haul trucking can’t rely on a high degree of accuracy of information, there will be a need to rely on more sensors to map the environment and weather conditions around the truck on the fly.

Elements of economics

When considering the economics of autonomous trucks, Salem reminds us that there are two elements of economics to consider when people talk about the vehicles. They are about gas or drivers, as well as the maintenance aspect of the trucks. To maintain the vehicles there will be a need for specialist capabilities, which will be expensive – particularly with respect to electric vehicles.

“The challenges are related to the battery to allow a reliable and sustainable journey for long haulage,” he explains while commenting that the charging infrastructure isn’t there yet for EV trucks. More to the point, he says it has been proven that, for now at least, the maintenance costs are comparable to those of the traditional vehicles. So, it’s hard to get the value out of EV trucks in terms of operational efficiency. With EV cars fuel is saved but with EV trucks the vehicle maintenance remains costly because of the need for specialist human capital.

He also suggests the discussion about autonomous mobility is very much coming from an ecosystem level. It’s not just about trucks, it’s also about the integration of the trucks. He said: “With different systems to ensure that the package delivery is optimized.” This is what he argues can be achieved with automation. For now, he says the process is semi-automated because the journey from A-to-B involves humans. However, with autonomous mobility the entire process can be automated and aligned with the different process steps to ensure “the most the efficient and timely delivery, which is not possible when a human driver is involved”.

Preventing fuel crises

So, could autonomous trucks have prevented the fuel crisis by delivering gas to service stations in the UK? That’s possible but, then again, problems could also arise if the specialist maintenance skills are short, or whether there is insufficient integration within the entire logistics ecosystem.

Still, Salem agrees that autonomy is critical because it will provide key efficiencies because the costs of a driver are high in that industry. High costs can cause issues in terms of driver supply. This is why the investment in autonomous long-hail haulage has accelerated, beginning with platooning. The big automaker are also putting in massive investment in development and piloting long haul autonomous technologies to extend operational efficiencies.