The document argues that automation’s “central irony” is that systems built to remove the human operator instead create new, more demanding forms of human work—monitoring, diagnosis, and skilled take-over—especially during abnormal conditions. It distinguishes normal from abnormal/emergency operation: in process plants, “normal operation and shut-down may be automatic,” while start-up and abnormal conditions remain manual, and in automated on-line control “a human operator is expected to take-over in abnormal conditions.” This matters because automation can deskill routine oversight, leaving operators less capable when rare failures occur.
A key mechanism is cognitive degradation. When operators “ride on” skills they no longer use, manual control competence deteriorates over time, and long-term knowledge depends on frequency of experience and feedback; reduced real control experience weakens later generations’ readiness. Working memory is also affected: operators must act with “minimum information” and cannot build the predictive internal model needed for effective on-line decision-making.
The paper supports these claims with evidence on human vigilance. Classic “vigilance” studies (e.g., Mackworth, 1950) show effective attention cannot be sustained when “very little happens” for more than about half an hour, forcing reliance on automatic alarms. Yet alarms and displays introduce further problems: alarm proliferation (“flashing red lights”), and automation “camouflaging” failures by controlling against variable changes so trends appear only after systems are beyond control. Even when alarms exist, operators may face an “impossible task” of meta-level checking whether computer rules are being followed.
Finally, the document notes that computer aiding can improve performance under higher workload (e.g., Ephrath and Young, 1981; cockpit simulator results), but may reduce planning depth under autopilot. Simulator training can preserve skills, though “unknown faults cannot be simulated,” limiting coverage for extreme events.
Article timeline
Build this article’s timeline
Extract dated events, key actors, and open questions using the same timeline analysis available in Signal.
Lisanne Bainbridge, Bibby, Edwards, and Pergamon Press Ltd
🤖 AI prefill
📰What
The document argues that automation’s “central irony” is that systems built to remove the human operator instead create new, more demanding forms of human work—monitoring, diagnosis, and skilled take-over—especially during abnormal conditi…
🤖 AI prefill
📅WhenWhen did this happen, or what is the timeframe?
📍Where
Great Britain, Baden-Baden, and F.R.G.
🤖 AI prefill
💡Why
Centers on automation of industrial processes, human operator, and man-machine systems
🤖 AI prefill
Login to answer and earn XP for each gap you fill.
Automation Paradoxautomation, paradox, irony, human operator, control systemsManual Takeover Skillstake-over, manual control, oscillation, gain, timingOperator Knowledge Limitscognitive skills, long-term memory, working storage, retrieval, trainingMonitoring and Vigilancemonitoring, vigilance, alarm systems, attention, steady-stateHuman Computer Collaborationhuman-computer collaboration, on-line decision-making, process industries, flight-deck automation, abnormal conditions
Key Phrases:automation of industrial processes98%human operator97%man-machine systems93%process control92%on-line control90%manual take-over89%abnormal conditions87%human-computer collaboration86%control engineering84%system failure and recovery83%
Manipulation Score: LowHow much this article uses emotional framing vs. straight reporting
Loaded Language225%
Absolutist615%
Threat/Panic125%
Us vs Them015%
Engagement Bait020%
Manipulation score look right?
Factual Claims (10)
mediumtechnology
Automation of industrial processes may expand rather than eliminate problems with the human operator.
mediumtechnology
Even highly automated systems, such as electric power networks, still need human beings for supervision, adjustment, maintenance, expansion, and improvement.
mediumtechnology
The more advanced a control system is, the more crucial the human operator's contribution may be.
lowtechnology
In process plants, different modes of operation may be automated to different extents, such as normal operation and shut-down being automatic while start-up and abnormal conditions are manual.
mediumtechnology
Experienced operators make smoother and quicker step changes than inexperienced operators.
mediumtechnology
Physical control skills deteriorate when they are not used, especially gain and timing refinements.
mediumtechnology
An operator who has been monitoring an automated process may become effectively inexperienced if asked to take over later.
lowscientific
Efficient retrieval of long-term knowledge depends on frequency of use.
mediumscientific
It is humanly impossible to maintain effective visual attention to a mostly inactive source for more than about half an hour.
mediumtechnology
Automatic alarm systems are needed to detect unlikely abnormalities because humans cannot reliably monitor them continuously.
Comments
Login to view and post comments.