The seven things we score before telling you where to start

Last month we put a number on what automation is worth on the airfield. About 400 automated vehicles can do the work of 1,000 manual tugs, and roughly $149 million in annual savings sits at the end of that transition.

A lot of you wrote back with the same question. Not whether it is worth it. You already believe that. The question was how do I know when my airport is ready to start.

So this month we answer it. These are the seven things we score before we tell any airport where to begin.

The pilots worked. The next stage is a different job.

The last few years of trials did exactly what they were supposed to do. They proved the vehicles work, some of them anyway. Tugs move bags. Mowers cut grass. Robots find debris on the airfield. That question is closed.

What comes next is a different job. Taking a vehicle from a supervised trial to something that runs every day is not more technology. It is a set of questions answered in writing. How do we know it is safe? What could go wrong, and what are we doing about each one? Where exactly is it allowed to operate? Who has the power to shut it down? That is integration work, and it usually takes different people than the trial did.

You do not buy that from a vendor. You build it. Almost nobody has built it before buying the hardware. The airport that does it in that order will move faster than everyone still doing it in reverse.

The seven

We score every airport across seven categories. Fourteen questions, each answered on a five-step scale, producing an overall score and a score for each category. Together they are the Airfield Autonomy Framework.

1. Foundation. Is there a real problem, or just interest in the technology?

Who is backing this at the top, and how clearly can you state the problem? The strongest answer is a specific pain with a dollar figure attached. The weakest is that the technology looks interesting.

This sits first for a reason. If you cannot say what problem you are solving, you cannot say what success looks like. And a trial with no definition of success produces goodwill instead of a decision. Decide what it worked means before anyone drives anything.

2. Operations. Who owns this on a Tuesday?

Not who champions it. Who owns it. Innovation teams start programs and operations teams run them, and the handoff between those two is the part worth planning early.

Then the harder half. When the vehicle does something unexpected, who has the authority to stop it, and who has the authority to start it back up? Those are rarely the same person, and both need names before you need them. Stopping should be instant, no permission required from anyone. Restarting should take a defined set of yeses.

3. Safety. Does your safety program cover a vehicle with no driver?

Most airports run a serious safety program. Almost none have extended it to a vehicle with nobody in the seat. A driverless vehicle outside your safety program is an asset nobody owns.

One thing worth knowing before your first risk review. Every risk gets judged on two things: how bad the worst outcome could be, and how likely it is. Your safeguards change the odds. They do not change the worst case. Teams new to this always try to argue the worst case down by pointing at their safeguards. It does not work that way, and knowing that in advance saves an afternoon.

4. Organization. Do the people who will simply encounter it know anything about it?

Getting everyone the vehicle touches on the same page, and preparing your workforce. The part almost everyone misses is that awareness is a separate job from training.

The team operating the vehicle gets trained. The ramp crews, the tenants, the fire service, and the handlers who merely work around it need something different and smaller: what it is, how to behave around it, and how to raise a concern. Leave them out and they will make their own decisions about it, including moving it out of their way.

5. Technology. It is more than the vehicle.

The vehicle needs a supporting cast. Reliable wireless coverage across your airfield, and connections into the systems you already run, so you get live monitoring instead of a dot on a map.

Then the question nobody asks until it bites. What does your construction schedule do to the routes you approved? Construction, detours, resurfacing and changed traffic patterns all mean taking a fresh look at where the vehicle is allowed to go, and every airfield has them constantly. The map and boundaries the vehicle runs on get rechecked after every change, not once at kickoff.

6. Planning. Is there a roadmap, or just a date?

The business case and the timeline. A date becomes useful the moment something is sequenced against it. On its own, it is a wish with a deadline.

And a scoring rule that matters more than it sounds. When you compare your options, never average the scores. A great cost score can hide a bad safety score, and that is exactly the mistake the exercise exists to catch. Every category has to pass on its own.

7. Decision. Could you evaluate a vendor today?

How you would buy it, and how you would judge it. This is where the two most important documents live.

The first is yours: a written description of where you expect this vehicle to operate. Weather, lighting, surfaces, traffic, radio coverage, all of it. The second is the vendor's: a written description of what their vehicle was built to handle. Evaluating a vendor is checking one list against the other, and you cannot do that until yours is written down. If you hear the acronyms TOD and ODD, that is all they mean. Your list, and theirs.

Most airports have never written one, and there was never a reason to. A driver holds all of it in their head and adjusts without being told. Write it down now, because the vehicle cannot figure out any of it on its own.

Then the last gate. Before you accept the system, decide what proof you require and who signs their name to it. Our rule is blunt. A blank line or a TBD means not accepted, and a claim without evidence does not count as proof.

What the scores usually say

The scale runs Aware, Exploring, Planning, Pilot-Ready, Deployment-Ready.

Most airports land in Aware or Exploring, and that is what you would expect. This is new work. Nothing until now required a safety plan for a driverless vehicle, or a written map of where one can operate, or proof standards for a vehicle you have not bought yet.

So an early score is not a verdict. It is a starting position, and almost everyone is standing in the same place.

What the score gives you is order. Which two or three categories to close before you spend real money on hardware, and in what sequence. That sequence is the whole value. Close them out of order and you do the same work twice.

The bottom line for airport leaders

You do not need a perfect score to start. You need to know where you sit and have a sequence for moving up.

The airports that will be ready when the economics turn are the ones building this foundation now, quietly, while the spreadsheet still says wait. The foundation takes longer to build than the vehicles take to buy.

This edition also appears on LinkedIn. Read and subscribe there.