Wiklund’s Perspective: Automation is (Mostly) a Good Thing

People who like cars (that includes me) lament the dramatic decline in the number of new cars offered with a manual transmission. I’ve always driven a car that requires you work a clutch and “row” through the gears. I think it is fun and provides me with what I consider an extra degree of control.

However, in the US, manual transmission cars have become quite rare. Even iconic sports cars like the latest Chevrolet Corvette are now exclusively equipped with automatics. The shift to automatics is even picking up in countries historically devoted to their manuals. In a short time, manual transmissions have effectively become anti-theft devices. Few young folks know how to drive them.

So, what is my point as it relates to working in human factors engineering (HFE) and specializing in medical technology? As technologies evolve, there is a tendency for manual functions to be automated in pursuit of greater ease-of-use and convenience. For me, it feels nostalgic to roll down a window by rotating a crank, but I don’t miss it. I also don’t miss “dialing” a telephone, particularly because calling home required me to dial a 10-digit number with several nines. The extra time required to dial nines was trivial but nonetheless annoying. The “newfangled” touch tone phone was an exciting innovation and made less work, if I can even call it work.

From an HFE standpoint, the main benefit of automation is that it largely eliminates the opportunity for a use error. Accordingly, automation is an excellent way to reduce use-related risk. No user task, no use error.

For example, a digital thermometer that requires a swipe across the forehead saves (1) literally shaking-down the thermometer to a low reading (i.e., physically driving the mercury down the tube to a low starting point), (2) placing it in the feverish person’s mouth or bottom for a minute or longer, and then (3) reading the level of mercury in the glass tube. The added steps were added opportunity for potentially consequential use error.

When introducing automation, the challenge is to make sure it does not create other opportunities for use error. One problem that can arise is the loss of situational awareness. That is, a user loses a sense for how a product/machine/system operates and is either less able or unable to intervene effectively when something goes wrong. This has become a too common problem in aviation, leaving conventionally trained pilots uncertain about what has gone wrong with their airplane and unable to respond correctly in order to avoid an accident. You may recall that this was a root cause of accidents involving the Boeing 737 Max 8 in 2019 and 2020. It was part of the problem decades ago when the reactor operators at the Three Mile Island nuclear power plant were unable to tell if the plant’s reactor was receiving enough cooling water. It has been part of the problem when people have caused harm using medical devices such as anesthesia machines, infusion pumps and dialysis machines, as well.

Based on an analysis of past accidents related to automation and loss of situational awareness, HFE specialists have drawn important lessons. One lesson is to find ways to keep the user in the loop so that she/he/they can respond to “upsets.” Practically speaking, this means providing users with adequate information and feedback so they can intervene in a process when necessary. The feedback might come in the form of an on-screen status indication.

A second lesson is to provide users with training (including hands-on practice) and troubleshooting guides so that they can take proper control of a process, when necessary, even when the need to do so is a rare event. This explains why there is often a quick reference card attached to medical devices, such as a ventilators and infusion pumps.

A third lesson is to provide a back-up, manual system in cases that the consequence of failing automation is severe. Therefore, hemodialysis machines have hand cranks to get blood back to a patient in the event the automated pumping mechanism fails.

Does this mean a car with electric windows and automatic transmission should also have a gear stick and window cranks to be safe? Well…I’ll say no on the gear stick for practical reasons and because there are things called tow trucks. But a power window failure in the event that a car becomes immersed in water is a concern. Maybe adding a back-up crank is also impractical but having a tool to break the window and escape a flooding car is practical. Every car should have one.

Oh, did I mention I just replaced my 13-year-old car with a newer, used car with an automatic?

Michael Wiklund is General Manager and Director, Human Factors Research & Design (HFR&D) at Emergo by UL.