Why Are We Still Training Humans to Operate Dirt Machines

B4DD9EA0 4190 4F68 A08C 798855133FF2 e1537628205948

Why Are We Still Training Humans to Operate Dirt Machines

I was watching some training videos this week featuring augmented and virtual reality, and concluded that the initiative and the technology were no longer on the right track.

Playing In The Dirt

When I was growing up, I always wanted to get behind the controls of some piece of earth moving equipment. My parents inadvertently nurtured this latent desire by feeding me my own fleet of miniature diggers, plows, lifts, loaders and cranes. Every Christmas, Santa unloaded yet another piece to the collection and I would impatiently wait out the winter so that by spring, I’d be out in the sandbox in the backyard, fending off the ants and tending my own earthen dikes to secure my dams. I conveniently overlooked the fact that the front end loader was at 1/64 scale, the hauler at 1/32, and I had to lift the entire loader off the ground to empty the bucket. 

My toys were pretty realistic—they had joints that work in a similar manner to the real thing, buckets rigged with pulley and string rather than hydraulics, and full steering systems or all-terrain treads.

It’s never occurred to me how the operators of this equipment actually learn their craft in the real world. How do you train on a front end loader, for example? Surely not on toys. A customer doesn’t want a trainee on their site making a muck of a trench job. The owner of the gear doesn’t want some newbie banging an expensive bucket loader into the bed of a dump truck (or tipper lorry). As this equipment gets larger (the biggest dump trucks hold up to 450 metric tons of material, and the loaders drop as much as 75 metric tons at a time), the consequences of errors escalate. 

Traditionally, training facilities consist of large indoor sheds over an earthen floor where students can dig away in climate controlled conditions until they master the basics (overall speed, dig accuracy, drop accuracy, bucket fill rate). Applying digital reality training solutions to this problem, students manipulate the joystick controls of a virtual digger to dig out virtual dirt for a virtual dump truck. Not surprisingly, the before and after results show the exact same gains as putting pilots into flight simulators—students learn faster, retain more, and perform better with fewer safety problems. Everyone feels good. 

Except we’re no longer solving the right problem. 

An Alternative Earthen Future

Have you been watching the developments at Rio Tinto? They’re clearly aiming to be the world leader in moving earth around. They operate really big mines in really awkward places extracting a broad assortment of rocks, ores, minerals and metals. They lead the deployment of the super sized autonomous heavy hauler in their mines, and this digital innovation is migrating to mines around the world. These massive robots operate without humans at the helm, including in the oil sands industry. 

And Rio doesn’t stop there. They operate the world’s first fully autonomous rail road, in far off Western Australia, that carries mined material from the inland mine to the port where the ore is off loaded to await ships to carry it to market. 

What’s next? I’m second-guessing Rio’s direction, but I can see a fully autonomous port that handles the loading of ships without humans manning the controls. Robot ports already exist, of course, for containers, but it’s not a stretch to visualise a bulk port that handles materials autonomously. Ghost ships (robot ocean-going vessels) are planned to ply the global trade routes.

Get the picture? A fully autonomous resource extraction robot or robots. Resource extraction will soon look more like a car plant, with industrial robots handling the work, from end to end. Why should the auto industry have all the fun? 

You might ask why Rio is working to take the human workforce out of their operations.  It’s pretty simple, really. Humans are expensive, unreliable and buggy. They need climate controlled operating conditions (but mines are in hot, wet, rugged places). Humans need to be protected from drops, trips, vapors, heat, explosions, sharps and chemicals (but mines have all of these features). Human concentration degrades over time (but mines need to run round the clock to be economic). Humans take breaks (but the assets only make money when they’re working). They trial dangerous shortcuts without thinking about the consequences (and bad things happen). People are prone to bouts of intoxication and revelry (neither is welcome in a mine).

Increasingly humans don’t want to work in hot, dangerous and demanding locations, and the few willing to do this work want a lot of money for their efforts.  

Earth First (We’ll Mine the Other Planets Later)

Watch the press, and you’ll spot the occasional story about some entrepreneur wanting to mine asteroids. Watch Avatar, and you’d think that the answer is having lots of handsome people running the show. That’s not the future I see unfolding. I see Rio’s robots handling the digging and the moving of earth around, from loader to hauler to conveyor to plant. We will send robots to do our bidding (and we already do—all those interplanetary probes and satellites we fire into space are rudimentary robots). And we will supervise the robots.

The big mining concerns are setting a new course. Rather than making work safer for humans, mining is simply changing the work to remove humans entirely from harms’ way, and setting the stage for a much bigger intergalactic future. 

Compare this to oil and gas companies and their suppliers. After they recruit some talent, companies then spend a small fortune trying to keep their human charges from hurting themselves. The amount of time and money being spent on improving safety, training for safety, measuring safety is staggering. There are legions of safety supervisors, sellers of safety equipment, standards bodies for safety. It’s even said in the industry that the way to get a project funded is to claim that it will improve safety or reliability, and as long as the argument is plausible, it gets a look. 

It’s like we’re in a safety arms race that no one can win and no one wants to pay for (try tacking 5% onto your bill for your safety program and getting your customer to pay for it. Good luck). And the industry is captive to a narrative that keeps this race going at full tilt.

Time For a New Business Model

Entrepreneurs are building robots that can fold laundry. It’s devilishly hard to fold clothes. Robots need to distinguish between trousers and tees, fronts from backs, widths from lengths, bell bottoms from capris. Humans do it with ease, but robots struggle. So how do you teach a robot to fold laundry? One way is to have an experienced laundry folder operate a set of controls that manipulates a robot’s arms, and by repeating the same movements over and over, with different clothing, the robot eventually learns the job. Human led machine learning. 

As you can see, the training simulator in the video above has it backwards now. It teaches humans how to manipulate the arms of the machine, rather than having humans teach the machine how to dig for itself. That’s how Tesla’s cars work. The cars today may be driven by humans, but the humans are also training the cars to take over the driving one day.

Don’t get me wrong. I like humans. I’m one of them. We’re creative problem solvers. We’re fantastic conversationalists. We’re funny and great company at parties. 

But we need to ask if our obsession with trying to make dangerous work safe from all possible harm is truly worthwhile because all harms cannot be fully eliminated and it’s costing the industry a fortune. At least one mining concern is now headed in a fundamentally different direction. 

The next generation workforce, who are much more digitally savvy, will struggle to understand why, in an era where robots make food, fly about, drive around, conduct surgery, patrol oceans and fold laundry, we still have back hoes with padded furniture. They will spare no thought to the argument that robots take jobs away. Their experience will be that robots actually create better jobs for people.

What Should Be Happening

If I were running an oil and gas outfit, I would set for the business a new North Star heading, one that challenges and overturns this business orthodoxy, that only humans can operate in the uncertain and demanding settings typical of extractive industries. I would invest instead in a series of steps that progressively move my company towards a different future. I would hang out with those organisations devoted to inventing the future, rather than with those intent on preserving the past. I would raise the acumen of my organization to deal with new technologies, like the internet of things, artificial intelligence and robotics.


If you own and operate an oil and gas business that routinely puts humans into harms’ way, and you have to compensate with lots of safety programming, I am worried for your future. The extractive industry agenda to try to make dangerous work safer appears to have run its course. The leaders are on a new path. There is a clear risk that you may be spending fruitless dollars on an endless pursuit while others are setting out to disrupt you. 


Mobile: ☎️+1(587)830-6900
email: ✉️ geoff@geoffreycann.com
Visit: 🖥 https://geoffreycann.com/
LinkedIn: 🚀 https://www.linkedin.com/in/digitalstrategyoilgas

No Comments

Post A Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.