Mentors on demand

Imagine if we could capture the knowledge from the industry's very best operators and make that knowledge available to newcomers. That idea is not as crazy as it sounds.

Remember when you were studying maths at school? The teacher was never satisfied that you’d got the correct answer without cheating. They insisted that you “show your workings.”

As a writer, I don’t normally do that; often because there are no workings; just a flash of inspiration and some furious typing.

But today, I’m going to make an exception. The story I’m about to lay before you pulls together several familiar threads: mentorship, driver aids, artificial intelligence, autonomous equipment, and the construction sites of the future.

I’m not going to address these factors in alphabetical or chronological order. Instead, I’ll explore them all as we go along, and do my best to weave them into a single, simple idea.

So why are we looking increasingly to autonomous equipment and artificial intelligence in the first place?

The answer is simple, yet complex. Straightforward, yet nuanced.

At heart, the main driver for autonomy is often said to be a growing shortage of skilled and competent operators. As an industry, we’ve stood idly by while our workforce has aged, and we’ve done precious little about it. In fact, at the same time we failed to attract sufficient young people to replace those at or near the end of their careers, we encouraged some of our most experienced operators toward the exit by removing their Grandfather Rights.

Of course, the industry's push for autonomy isn’t solely due to the shortage of skilled operators.

Humans are fallible, vulnerable and, therefore, unreliable. They expect holidays. They are susceptible to sickness (real or imagined). Some show up drunk or high. Some don’t show up at all.

Worse still, human decisions are often driven by factors other than logic. Some operators crank the revs up to full throttle because they like the roar of the engine, with no thought for fuel consumption. Some prioritise speed over safety. Some cut corners simply because it’s Friday afternoon and the pub is calling to them.

Autonomous machines, by contrast, are driven by pure, unassailable logic. They do what they’re programmed to do and they don’t deviate. They extract every last ounce of value from their fuel and are perfect custodians of themselves, constantly monitoring for faults or wear.

It’s an uncomfortable truth; but autonomous machines are faster, more efficient, and more economical than their human counterparts. And they don’t get hangovers.

So that’s where we are heading. But the journey from fully human to fully autonomous is not a straight line. There will be stop-offs and diversions along the way. One of those stop-offs is the halfway house of driver aids.

At the most basic level, driver aids act as constraints, preventing over-digging or over-extending. But now we’re seeing the emergence of a new generation of aids that enhance, not restrict; that complement rather than control.

A prime example is the Intelligent Machine Control built into the new Komatsu PC220LCi-12. This system introduces auto swing, enabling semi-automatic truck loading. In simple terms, it records an operator’s dig, boom raise, and dump actions, and then repeats them with virtually no human input .

Now, keep that “record-and-replicate” capability in mind while I take a slight tangent.

Modern machines are bursting with software systems; and those systems open the door to entirely new business models.

Bobcat, for example, has floated the idea of paying to unlock additional power capabilities for particularly tough applications. Caterpillar, too, has apparently pre-installed functions in some machines that could be activated through licensing or rental.

This is where we begin to tie together all those threads - AI, autonomy, and future job sites - with a big bow marked mentorship.

It is possible that the skills of our very best operators could live on in software form.

We already know that Komatsu can replicate a human operator’s actions. We know Bobcat and Caterpillar envision a future of software-unlocked functionality.

So how about this as an idea:

Imagine if we could digitally capture the skills of an experienced operator; skills that could then be licensed, rented, or unlocked depending on the task.

You could unlock the “Gary” module for super-efficient truck loading. The “Jim” module for precise trenching. The “Dan” module for deep drainage. Or the “Paul” module for high-reach demolition.

A fusion of AI and intelligence capture could bridge the skills gap between today’s human-led operations and tomorrow’s semi or fully-automated job sites.

We might not have mentors in the traditional sense. But it is possible that the skills of our very best operators could live on in software form. And this could, conceivably, extend way beyond mere mentorship.

Imagine a rookie operator stepping into the cab of a modern excavator. Instead of spending months shadowing a seasoned veteran, they activate a digital module; maybe the "Gary" module for efficient truck loading. Instantly, they're guided by AI-driven prompts, replicating the refined techniques of an expert.

This scenario isn't a distant dream; it's an emerging reality with the ability to reshape how we approach training and daily operations in the construction industry.

Traditionally, mastering heavy machinery required prolonged apprenticeships, with novices learning through observation and gradual hands-on experience. With the advent of AI and machine learning, new operators could access on-demand training modules that encapsulate the best practices of seasoned professionals. These modules could offer real-time guidance, allowing learners to practice and internalise optimal techniques from the outset.

The integration of AI doesn't stop at initial training. Suppose an operator encounters challenges with trenching stability mid-project. The system can promptly suggest activating the "Jim" module, providing immediate, task-specific guidance. Such micro-learning interventions would enable operators to address issues in real-time, minimising delays and maintaining project momentum.

Access to top-tier expertise has traditionally been limited to larger firms with the resources to hire and retain seasoned operators could unlock specialised skills as needed, ensuring high-quality performance without the overhead of full-time specialists.

This democratisation would not only enhance the capabilities of individual operators but also elevate the industry's overall competency, fostering a more competitive and proficient workforce.

The flexibility offered by these modules allows for dynamic allocation of skills across various tasks. For instance, an operator might utilise the "Dan" module for deep drainage in the morning and switch to the "Gary" module for truck loading in the afternoon. This adaptability could potentially reduce reliance on specialised personnel and enable teams to respond swiftly to changing project demands.

We are currently losing experienced operators far too quickly; and we’re attracting newcomers far too slowly. Against this background, we are in very serious danger of squandering the knowledge and wisdom of some of our best before it can be passed on to the next generation.

But by harnessing AI technology, those newcomers could still draw upon all that accumulated knowledge, long after those industry veterans have hung up their boots.