An Initial Model for Autonomous Trucks in Australia?

Updated with long distance vehicle announcements


A recent announcement in the United Kingdom has the government allocating 8.1 million pounds to a truck platooning trial:

Semi-automated truck convoys get green light for UK trials

Platooning is essentially like bicycle pelotons in road races like the Tour de France, where riders get sucked along in the slipstream. Until you have actually participated in one, you do not realise how much easier it is to ride in the group. I knew that intellectually, but the experience is something else. For trucks this means less congestion and less fuel use. In order to achieve these results the artificial intelligence and sensing systems that controls the trucks have to be much better than human drivers so that the trucks can drive closer together.  In the UK trial the speeds and steering will be controlled by the lead vehicle.

Total autonomy for vehicles on the road is known in the industry as Level 5 autonomy. This is where vehicles can control themselves in all road conditions. We are a long away from this technologically, so the trucks in the trials will have human drivers who can take the wheel at any time. The problem with this is that driver attention will naturally wane and this may impact on reaction time. In this trial this may be dealt with by periodic blocks of time where the human driver must take command of the truck – whether there is a need or not.

The medium term adoption pathway here in Australia may be different due to the road conditions and distances travelled. Here in Australia the situation for truck driving is a little different than the UK. There are much larger travel distances between the major cities, and the major inter-capital highways are less crowded. This is mirrored in the United States, especially in larger states such as Texas and California. This means that the adoption process of the technology may be significantly different.

There are a couple of technology issues in the adoption pathway that is chosen that flow into these sorts of differences and how we might choose to adopt the technologies.

Firstly there is a significant debate in the autonomous vehicle technology world about the approach of using maps versus continuous sensing. As humans we can navigate an unfamiliar terrain because our sensing and vision systems are good enough to recognise and continually process information at a level that is useful. The technology in autonomous vehicles is still not good enough to achieve that yet, and this is where mapping comes in. If an autonomous vehicle has stored in its system a map of the territory it is about to navigate, it only has to compare the environment it is encountering versus the map. This significantly reduces the job that needs to be done, reducing the pressure on the technology. In the long run it is likely that onboard vision and sense making systems will be good enough to do without maps. In the short term  having maps significantly improves performance. The timing of these changes, and the implications for strategic competitive advantage are critical when thinking about strategic decisions for individual companies, and what the overall outcomes might look like (see: Winner-takes all effects in autonomous cars for an excellent discussion on this).

Secondly, at what point will we be comfortable with no driver in the vehicle, and will this be at Level 4 or Level 5 autonomy. At Level 4 autonomy the vehicle can drive itself but is limited either by geography or conditions. This means that while the driver can be removed there needs to be some sort of geofencing, or emergency failsafe systems. For example trucks on the highway may automatically pull over if rain levels go beyond a certain level, affecting visibility. If adoption pathways can be achieved at level 4 rather than level 5 then adoption will occur more rapidly as the technology will not have to be as advanced to achieve the outcome.

So if we can build a model in a specific area of trucking where there are less complicated driving challenges, and mapping  makes a significant contribution we can create faster adoption. Which takes us back to the highways between capital cities in Australia.

In Australia 18-19% of total road freight movements are inter-capital freight movements (Truck Industry Fleet Report 2015), and there has been significant improvement in those roads over the last 20 years. For example once we get outside of the major urban areas of Melbourne and  Sydney the road between the two cities is excellent for trucks. An early adoption model for autonomous truck movements in Australia might start with transfers between Melbourne and Sydney and look like the following:

  1. Autonomous trucks operating the full distance between the two cities except for the last 30 kilometres (plus or minus) in each city.
  2. A truck changeover system on the outskirts of both cities where either the truck takes on a driver, or the prime mover is changed over to a non autonomous prime mover and driver. This is necessary in an early adoption model because the challenges of driving in the major cities are significantly higher than on the open highway.
  3. A cooperative mapping effort coordinated by the Federal Government where the road is mapped in its entirety.
  4. The formal mapping is supplemented by all autonomous trucks contributing their mapping and sensing data to a central system to continually update the maps. Therefore any new hazards or changes such as roadworks are rapidly incorporated into the maps that all autonomous trucks use.
  5. Autonomous truck support centres where the control of the truck can be taken over by a remote driver in the case of difficulties such as problems with sensors, or road conditions which are outside of specified parameters.

Many of the pieces of such an implementation pathway are already in place or soon will be. Autonomous trucks have been trialled in several locations around the world, and we already have remote control of mining systems (Mining industry looks towards a new wave of automation ,  Rio Tinto: rolling out the world’s first fully driverless mines ). We also have remote control of drones for military operations.

Around the world the trucking industry is seeing problems with an ageing workforce, with trucking jobs being seen as unattractive by younger generations (Wheels not in motion: Australia running short of truckies). A system as described above can solve some of this problem by:

  1. Autonomous trucks can operate for more hours than human drivers can, increasing efficiency of truck use and reducing overall demand for drivers.
  2. Increasing the attractiveness of trucking jobs. In many cases the long hours and time away from home are significant factors reducing the attractiveness of driving a truck. If the long distances can be handled by autonomous trucks, and the drivers can go home to their families at night then the job becomes more attractive.
  3. A truck driving job is more interesting, as the easy parts are taken over by autonomous trucks, and the more difficult driving conditions, unloading operations, and interactions with customers are covered by human drivers in short haul operations.

Eventually most trucking operations will be carried out by autonomous trucks If we want to address the shortage of current workers, reduce fuel consumption for long haul freight, and possibly reduce fatigue related accidents, a model which accelerates early adoption should be trialled.


Proterra has announced an 1100 mile (1772.2km) trip of its Catalyst Bus on a single charge. (Proterra Counters Tesla’s ‘Beast’ Of A Semi With 1,100-Mile Range Electric Bus). In addition Tesla will announce its new Semi truck in October. With distances between Melbourne and Sydney of approximately 865 km, Sydney to Brisbane of 928 km, and Melbourne to Adelaide of 725 km this seems to put the intercapital freight market in the sights of autonomous electric trucks.

I am writing a book on autonomous vehicles with Dr Chris Rice of the University of Texas Austin. It is called Rise of the Autobots: How Driverless Vehicles will Transform our Economies and our Communities. Stay tuned for more excerpts as we finalise the book.


Note: The featured image comes from: 











Sell Your Crash Repair Business Now*

*this should not be taken as financial or business advice. If you own a crash repair business please take professional advice before making any decisions.

I am just going through the process of getting some minor damage repaired on our car and have been ruminating on the future of the insurance and repair model when we have driverless (autonomous) cars. This was also prompted by a couple of stories in The Age here in Melbourne:

Crash repair: How Ray Malone became head of ASX-listed company AMA Group


Driverless vehicles technology to roll out on the Tulla under trial


The first story describes how Ray Malone has built a Australia’s largest crash repair business, and is aiming to grow it even further. That would seem to go against the title of this post but it actually feeds into my thinking because Ray’s company provides wholesale service aftercare which will be vital in the scenario I am describing.

The second story is about how trials of driverless cars are starting here in Melbourne. This follows a large number of trials that are being conducted in various countries around the world.

Once we move to a reasonably widespread adoption of autonomous/driverless cars the local crash repair business will basically disappear except for a few large operators like Ray Malone but even his business could be under threat . The key reasons for this are:

1/ It has been forecast that autonomous cars will significantly reduce the number of car accidents that occur. This is based largely on the statistics that human error causes more than 90% of traffic crashes. So if we can eliminate the crashes caused by idiots, people under the influence of drugs and alcohol, and people driving tired or angry (Police looked into the deaths of 86 people on Victorian roads last year and found that in more than 10 per cent of cases the driver had experienced a traumatic or upsetting event.) we can significantly reduce the number of accidents.

Against this argument is that autonomous cars supplying a transport service may result in people travelling further and perhaps take more risks. Certainly it will allow elderly people who cannot drive, and young people who do not have a licence to travel in cars more than they otherwise would. There have also been arguments that because we feel safer we may take more risks as pedestrians or cyclists.  If we are conservative and say that only 50% of accidents caused by human behaviour will be eliminated we still have a significant fall in accidents.

2/ It is highly likely that we will see large fleet models emerge where large numbers of people choose not to own a vehicle. If the overall travel costs are lower than owning your own vehicle, and you can get a vehicle anytime you need one then the convenience of transport as a service outweighs the personal ownership model.  The economics for fleet owners are different than for individual owners when it comes to crash repair services. Fleet owners will want large scale service operations to reduce costs or will pay far less for the services of smaller scale operators. This feeds into a large supplier (such as Ray Malone’s company) snapping up more business. Larger scale crash repair businesses will benefit from the economies of scale that allow them to use new technologies such as robotics to increase throughput and reduce costs.

3/ The model for crash repair business location will change. Currently crash repair businesses are located in scattered locations throughout the suburbs and inner city. This is because if I want to take my car in for crash repairs there is a significant time cost for me to take my car to a location that is not near to my house or business. I have to travel to the crash repair business, and then get back to my home or place of work. So I want the crash repair business to be reasonably close. The location is mainly driven by the customer. If my personal driverless car needs crash repairs it can drive itself to the crash repair site, and a fleet service or a shared personal car service can replace my transport needs in the meantime.

If I was asked to drive my car (actual damage pictured below) to a service centre 40 km away I would not be very happy, but if my car can take itself then location becomes much less important and the costs of the business become far more important. Locating the crash repair business in areas of lower property costs with good transport links makes far more sense. It also means that the employees of the business will have lower property costs if they live locally. We already see this model in light manufacturing and food processing/handling facilities locating around hubs on ring roads, away from  inner suburbs with high property prices.

If a fleet ownership model predominates over personal ownership this effect will be even higher as large scale fleets look for cost reductions through economies of scale.

corolla damage 1


So if we summarise all the factors together if we assume a 45% reduction in total accidents (50% of human error crashes) and a tripling of scale that comes from the changes described above we get an 82% reduction in the number of crash repair businesses in any city.  I believe that the changes in scale may be even higher and we may end up with only 5-10% of the number of current crash repair businesses being economically viable.

If I own a crash repair business in any suburb in any of our major cities I will come under pressure from a high scale panel beater business set up on the fringes of the city with lower property costs.

So, if you are a crash repair business:

  1. Assess whether now is a good time to sell to someone else who does not understand these changes.
  2. If you think I am wrong then you should suspend that thought for just a few minutes and  think about what it means to your business and your assets if I am right. Even if you think that chance is only 5% you should set up a series of questions for yourself to monitor in coming years so that you can change your mind if the changes start to happen. Those questions include:
  • Is the practical outcome of accident reduction matching the rhetoric of the technology experts and the modellers? Look for signs of early change, cities where adoption is at the forefront of the change and make an assessment as to whether the predictions on accident reduction are true (or even going to be exceeded) and then think about the timing of the implications.
  • Look for areas or cities where the first full scale mass adoption of driverless cars might take place. For example Singapore, with a small land mass, and a relatively authoritarian government might be one. This will give you early signs of what larger scale adoption might look like.
  • Is the adoption model going to be a personal one or a mass fleet one? If the model is primarily a personal one then you should be thinking about whether you can become one of the new mega panel beaters on the fringes of the city that will survive the change. If the model looks to be a primarily mass fleet adoption one then there are less possibilities. Those fleet operators will either run their own operations which are standardised and mechanised or they will use their economies of scale to drive down margins in the businesses that supply them. You can still run a good business that way but the opportunities will be limited and will require lots of capital to create the volume throughput and economies of scale required. You will have to compete with the Ray Malone’s of this world.
  • Are any early models of very large scale, city fringe located crash repair businesses starting to emerge anywhere around the world? Are they successful?
  • Are car companies changing their business models for car repairs. For instance electric cars have far less moving parts than internal combustion cars. Does that make a difference to your business model? Are modularised car construction and repair systems emerging that will increase the capacity to adopt robotic repair and maintenance systems that will advantage large throughput car repair and maintenance systems?

While these changes may take 15 years to start to significantly impact on the crash repair business, once they become obvious the window to realise the business value by sale will quickly snap shut.

This is just one of the many implications of change from the widescale adoption of driverless cars.I am writing a book on driverless vehicles with Chris Rice (@ricetopher). It is called “Rise of the Autobots: How driverless vehicles will transform our economies and our communities. Stay tuned for more writing as we develop our thinking further.


Paul Higgins

The Supermarkets Demise – A Scenario

Back in November I wrote a post entitled: Are The Two Major Supermarkets in Australia Doomed?

If you are at all involved in the retail food chain I suggest you go and read it in full. The short answer is yes, but it will be a slow train crash.

A story in MIT Technology Review last week illustrates one of the possible models that can replace the supermarket model of today:

Autonomous Grocery Vans Are Making Deliveries in London


Of course supermarkets will be trying to incorporate such systems into their business model as well but my view is that because of their underlying legacy systems they will find the transition close to impossible.

The story is about a quite limited trial but it points towards a possible future:

“On the back of the vehicle are eight pods, each with a crate that can hold three bags of groceries. The van is filled by human hands from a small distribution center—in this case, a larger Ocado van, which stores 80 of those crates—and sets off following a route to its drop-offs, which is broadly planned in the cloud but ultimately executed by the vehicle. When it arrives at an address, the customer is alerted via smartphone and must press a button on the vehicle to open a pod door and grab the groceries.”

In terms of the final use case:

“Clarke imagines vehicles like these being used to provide on-demand delivery of groceries from a small nearby distribution hub, so that instead of booking a delivery slot customers hail their groceries—when they arrive home from work, say, even if it’s late at night.”

This ties in with an interesting analysis of the IPO for Blue Apron, the food company which delivers meal recipes and the main ingredients for those meals to your door. In that analysis in the New York Times, chef Amanda Cohen theorised that the Blue Apron model may destroy itself. She describes the fact (which went against her initial view) that many people she has spoken to said that the Blue Apron process had given them the confidence to cook more. If she is correct then this means that Blue Apron is training its customers not to need it any more, not a great business model as it means lifetime value of a customer may be severely limited.

The combination of these stories may point to a completely different future. As Amanda Cohen says:

‘” In Hong Kong, many people swing by a “wet market” on their way home from work and pick up the vegetables, fish or beef they’re going to eat that night. Same thing in France, Latin America, South Korea or pretty much everywhere people don’t load up their giant S.U.V.s with giant quantities of groceries to store in their giant fridges once a week. The meal kit model of keeping some staples in the cupboard and getting the fresh stuff as you need it is the market way of doing things”

One of the major problems with food delivery systems and in particular with automated delivery systems is what do you do with the fresh stuff because timeliness and the refrigeration process really matters. This is exacerbated by the fact that people are home at different times of the day or night and cannot necessarily take delivery when the delivery system wants to deliver . Various ways of solving this have been proposed including smart delivery lockers in apartment buildings or the local post office, etc. I can see that models emerging where all of the non-fresh goods can be delivered by an automated delivery system from a small local storage facility where you request delivery when you are home, just like you do when requesting an Uber right now. There may even be discounts for people who take quick delivery so storage space is always available, or people who will take a shared delivery and therefore will wait longer.

If this is part of a wider adoption of driverless cars then it can be part of a larger change. Driverless cars do not need to park, or at least do not need to park in busy or congested areas. I am an advocate for a driverless car adoption model where government or privately owned fleets provide transport as a service and surpasses the personal vehicle ownership model that has dominated the last hundred years. Even if that does not come true individual owners can hire out their driverless car when they are not using it so it does not have to be parked in front of the house or the office, or at the train station.

I I imagine a changed urban environment where mass adoption of autonomous vehicles changes the urban landscape by freeing up parking areas on streets and parking facilities . The freed up space on streets creates the capacity for more foot traffic, and increases in safe bike lanes while, driverless vehicles increase the capacity for people to travel for short trips locally. The parking facilities can be repurposed for storage and/or specialty markets for fresh products.
In that changed local environment we could see a model where large scale supermarkets are no longer the norm, where specialty fresh food stores spring up everywhere within easy travel distance of people’s homes. These specialty stores would be powered by the back end logistics that Amazon creates for Whole Foods, or their competitors (go read Ben Thompson’s excellent post: AMAZON’S NEW CUSTOMER for more details on their strategy) You would pick up your fresh product and speciality items on your way home from work or by a short walk or bike ride, or driverless car ride to the local store. Automated vehicles would deliver the staples to your door on request using pre planned orders or automated ordering systems like the Amazon Dash Wand.

In many areas this could revive the concept of neighbourhoods that really work in urban environments.

There are many ways the supermarket model will be attacked in the future. This is just one possible scenario. Given the pace of driverless car adoption and capacity for the car industry to deliver the full model is still a fair way off. The automated delivery system is not so far off. It fits the four level of automated driving systems by being in a geofenced area (local delivery only from a small storage/transfer facility), and carried out at low speed to reduce the risk of accidents. Full level 5 driving automation where vehicles can go anywhere in all conditions and no driver actions required are a lot further off. That does not mean there will not be continuing experiments with automated food delivery systems.

As Ben Thompson states in his article groceries are about 20% of consumer spending (USA). That is a big prize and lots of people are going to be going after it. Long term an automated vehicle delivery system will be a part of that. How big a part, and in what form remains to be seen.


I am writing a book on the adoption of driverless cars with Chris Rice entitled Rise of the Autobots: How driverless vehicles will transform our societies and our economies. Follow me here or on Twitter for more updates as we write and publish.

Paul Higgins




Augmentation of Human Capacity

On Friday I did the opening keynote for the Mindshop Australia conference. The title was “Bringing the Future into your Advisory Practice”. The focus was on ways of creating more value for the clients of advisors in the network. After the session there was much discussion from participants on the nature of work and the sorts of jobs that they should encourage their children to be aiming for.

My response to those questions was to use examples to highlight principles rather than recommend specific jobs because jobs will change. I used the example of the health sector and new AI developments in my presentation as well as in the discussions afterwards. For example:

Self-taught artificial intelligence beats doctors at predicting heart attacks

stylised heart image from sciencemag

On the weekend I was then reading Stowe Boyd’s  10 work skills for the postnormal era and I was struck by the statement on “Freestyling” from Tyler Cohen:

“When humans team up with computers to play chess, the humans who do best are not necessarily the strongest players. They’re the ones who are modest, and who know when to listen to the computer. Often, what the human adds is knowledge of when the computer needs to look more deeply”

This married up with the response I was giving to participants at the conference. The use of AI systems to augment the capacities of humans  does not augment everyone equally. In the world of medical specialists it is a commonly held view among patients that they will put up with specialists with poor social skills or high prices because of the knowledge they hold (putting aside the issues of the professions restricting supply to keep prices high).

If that knowledge moves largely to the realm of artificial intelligence then this re-weights the value of the medical specialist. If the machine can do things the individual or team cannot possibly do by being able to access more knowledge and make more connections in that knowledge than is humanly possible then it changes the system. Knowledge becomes less important and skills such as the capacity to work with the AI, patient empathy and general social skills become more important.

Augmentation  of human cognitive capacities will do that across sectors and industries.


The Future Competitiveness of Corporate v Individual Medical and Veterinary Practice Models. Is AI the key?

Before Christmas I did some work on the future of veterinary surgeons and the education and regulatory changes that might have to occur to move with those changes. One of the things that occupied my mind with that work was the issue of artificial intelligence systems on the competitive position of veterinary practices. My belief is that artificial intelligence augmentation of medical/veterinary capabilities is coming quite quickly and a large scale corporate practice model has significant advantages in this space over the individual practice.

Increasingly we have seen a more large scale corporate model in the veterinary practice market in Australia and around the world. Examples in Australia include Green Cross for major city practices and Apiam Animal Health in rural practice.  This has followed similar trends in medical practices where you have organisations like Medical One and Tristar which started out in rural Australia but has expanded into the cities as well.

The basic business model and value proposition of the corporate model is:

  • Presenting a single branded product almost like McDonalds so a patient or client can be confident of going to the practice no  matter where they are.
  • Increasing purchasing power of all of the back end parts of the business from pathology services down to supplies of bandages, etc. This is more important in the veterinary model where practitioners are able to prescribe and supply vaccines and S4 drugs and make a profit from them.
  • Taking over the administrative and compliance parts of the business to allow practitioners to focus on patients. This can include the standardisation and delivery of training requirements, building management, OH&S requirements etc. It can also extend to the supply of consulting hours as a service package.
  • Networking and support for practitioners in smaller practices.
  • The spread of investment and risk over a larger geographic range and customer base.
  • A much more secure retirement/part time/business sale option for owners and practitioners. On a personal note my previous long time GP semi-retired by moving into a Medical One practice and then progressively handing over his clients to the other doctors in a very caring and professional manner.

These models have expanded quite significantly over the last decade which speaks to the financial viability of the model and while there has been the odd flare up and accusation of over-servicing in general the models seem to have worked. I originally started my working life as a general practice veterinarian and the corporatisation of veterinary practices is a hot topic at the reunions I have attended.

The competitive position of the individual practice has been built around personalised service and attempting to be portrayed as caring more for the animals they serve than the big corporate competitor. many of these practices are still doing quite well but the trend is towards the corporate practice.

Which brings us to future competitive positions. I believe that we are rapidly heading towards a future where augmentation of medical and veterinary skills via artificial intelligence is going to be  a ticket to play in the game. This is going to be a narrow focused intelligence rather than a general intelligence. In the medical space we are seeing story after story emerge of new models where AI systems are getting as good as human doctors or have an edge over human doctors:

This AI Can Diagnose a Rare Eye Condition as Well as a Human Doctor

eye image from motherboard vice artificial intelligence

Predicting non-small cell lung cancer prognosis by fully automated microscopic pathology image features

AI is nearly as good as humans in detecting breast cancer

Self-taught artificial intelligence beats doctors at predicting heart attacks

All of the examples above are based on variants of machine learning and one of the defining characteristic of machine learning artificial intelligence systems is they need large data sets. In the heart attack example above the artificial intelligence trained itself on almost 300,000 case records. As we currently understand artificial intelligence, that system is not transferable to cancer diagnosis, it remains a specilaised cardiac application. The non-small cancer system above was trained  on almost 3,000 images and subsequent patient follow up records.

There are two ways to get a large data set to train on in the medical/veterinary field. The first is to work on aggregated image and case records as in the examples above. That will certainly be a major part of the market. Large capital expenditures will be required to assemble the required images/case files and process them in a way that improves patient diagnosis and outcomes when used in conjunction with human doctors. So we will see services offered for specialised areas such as heart attack prediction, organ by organ cancer diagnosis, etc. As always the areas that have the largest and most affluent markets will be the services that are first offered.

However we are also moving in to a world where artificial intelligence systems can be harnessed by smaller players at much lower cost. Take this example of How a Japanese cucumber farmer is using deep learning and TensorFlow:

cucumber-farmer-14 tensor flow artificial intelligence

The son of a Japanese cucumber farmer (who admittedly had some very good tech skills) built an artificial intelligence based cucumber sorting system for his parents. The problem was training the system required a lot of images and the process took a lot of computing power. More and more in the future you will be able to plug into a cloud based machine learning platform that will enable you to harness much more computing power that is specialised to do this sort of a job for you.

So a large network of medical or veterinary practices could offer services that are not offered in the general market by collecting all the data on their patients across all their practices and using machine learning platforms to train a system that is specific to their patient database.

This will be impossible for the individual practice to match. Is this the killer application for the corporate model to almost completely squeeze out the individual practice?

If so I will be watching for models emerging in the veterinary space because the regulatory hurdles and insurance requirements are much lower. If they are successful then I see those sorts of models flowing into medical practices.