Autonomous Trucks: Where The Rubber Meets The Robot

It’s a brave new world. Artificial intelligence, decarbonisation and automation. Three technologies at the front lines of engineering advancement, pushing ever further into the unknown. Our industry may be standing on the precipice of major change. Diesel to digital. Conventional to calculated. Elbow grease to automated. Will the impact on the trucking industry be revolution, or merely evolution? In the first of a three-part series, The Tech’s Files unpacks vehicle automation technology, and makes the case for its best use in Australia’s trucking industry, and the critical path forward for success.

Driving Forces: What Is Autonomous Driving Technology?

Broadly, Autonomous Driving Systems (ADS) combine a range of technologies, mostly existing, but with some new features. These can be split up into two categories: hardware and software.

The hardware is easily to visualise; it comprises sensors, cameras, computers and actuators. The sensors and cameras monitor the driving environment, the computer can process the information they collect and the actuators can control the vehicle’s steering, braking and acceleration. In that sense, they all operate together in the same way a human does.

Daimler Truck recently unveiled a battery electric autonomous Freightliner eCascadia technology demonstrator

The software isn’t as easy to comprehend. It comprises a complex set of programs that may collectively have up to 300 million individual lines of computer code (for reference, a conventional vehicle may have up to 100 million). Automated driving system software may also feature advanced types of programs including machine learning and artificial intelligence, but more on that in future articles.

Depending on the vehicle and its level of sophistication, the hardware and software can be capable of performing the entire driving task, without any human input, for all or part of a journey, which in itself is nothing short of revolutionary.

Level Up: Varying Stages Of Sophistication

Levels of autonomous technology are often described according to the Society of Automotive Engineers (SAE) classifications, numbered 3, 4 and 5, with 5 being the most advanced level.

An ADS with level 3 automation (also called ‘conditional automation’) means that the system can perform every aspect of the dynamic driving task, but only in certain situations. A human is required to remain in the driver’s seat, commonly called the ‘fallback-ready user’. While they may not need to actively monitor the road, they must be ready to take over when required to by the system.

Volvo Autonomous Solutions offers a complete solution based on a Transport as a Service (TaaS) model

An ADS with level 4 automation has similar capabilities to level 3, but crucially does not require a fallback-ready user, as the vehicle must have the ability to always reach a safe state without requiring any intervention from a human driver. However, like level 3, the system may still only be capable of operating in certain situations, such as completing a defined route, or while travelling on highways.

An ADS with level 5 automation removes the situational constraints and conditions from level 4, in that the vehicle is capable of operating itself on every road environment, in all situations. A human is not required to be present in the vehicle at all. For those vehicles, not only has the rubber met the robot, but the two are inextricably linked.

Automatic For The People: Where Are We At?

Encouragingly, Australia is leading the way in the roll-out of automated trucks in the mining sector, with some reporting the size of the automated fleet at more than 500 units, while most other countries have only managed a handful each. The manufacturers involved include Scania, Volvo, Caterpillar, Komatsu and Hitachi.

It is important to note, however, that the operating environments of those vehicles are limited and highly controlled. Whilst automated technologies have been developing for decades, in terms of their on-road use, technology level 3 and above essentially remains within the ‘limited trial’ phase.

Australia is leading the way in the roll-out of automated trucks in the mining sector

To-date, there have been several public trials of automated trucks in Australia, most recently within the Port of Melbourne precinct and on the CityLink and Monash Freeways, via a partnership with Transurban, IVECO and Plus, a global autonomous driving software company. These tests have focussed on aspects of the technology required to reach Level 4 automation.

Despite those initiatives, the extent of practical experience is low, and the risks that may present themselves via more widespread adoption in Australia remains largely unknown.

AI Is My Co-Pilot: What Could An Automated Heavy Vehicle Do?

At this early stage, two of the most attractive uses of automated vehicle technology for heavy vehicles are in the areas of fatigue management and overcoming the perennial labour shortage issue.

ADS capable of level 3 automation, and those at higher levels, may provide heavy vehicle drivers with the opportunity to rest while the vehicle is in control. If this benefit is realised, it could boost productivity, alleviate parking issues at rest areas, and reduce driver fatigue. ADS capable of level 5 automation could remove the need for a human driver entirely.

A reduction in crashes caused by human error, reduced fuel usage through optimised driving patterns, and reduced wear and tear leading to fewer services and inspections may also be other possible outcomes.

Port of Melbourne, in partnership with Transurban, IVECO and Plus, has trialled automated trucks in Australia

So, is it going it happen? It is likely to hinge on the eventual laws that will govern ADS. The National Transport Commission (NTC) and the Department of Infrastructure, Transport, Regional Development, Communications and the Arts (DoITRDCA) has been conducting extensive work in this area since 2016, and recently concluded a consultation period that posed these questions to industry, alongside dozens of others of equal complexity.

Consider just some of these questions – is a fallback-ready user still performing a work-related task while sitting in the driver’s seat, even if they are not actively monitoring the road? In doing so, would they be as fatigued as a driver? If not, how much less fatigued might they be?

Steering Clear Of Past Mistakes

The future legislation, in whatever form it takes, should readily allow industry to use technology to deliver improved outcomes and overcome challenges, as opposed to creating new barriers or limiting the benefit to an unattractive level.

Our industry has witnessed first-hand the on-going problems that ineffectual legislation can cause. The primary example is the Heavy Vehicle National Law (HVNL), which is widely regarded as overly long, complicated, inefficient and too prescriptive.

Scania recently opened up for orders of the company’s self-driving mining solutions in Australia

Future national laws of similar magnitude and importance must expressly avoid repeating its mistakes and provide a fertile ground where the transport industry can make best use of all the safety, productivity and efficiency improvements that automated vehicle technology can eventually offer.

The degree to which it can achieve that, will determine if autonomous truck technology can redefine our industry, or merely refine it.

---

Coming Up Next: The Road To Decarbonising Australia’s Transport Industry

---