Skip to content
The Future Explorerby Benjamin Talin
Enablertechnological

Automation

Machines doing tasks that people used to do, from looms and assembly-line robots to software that files paperwork. It is the oldest enabler on the map, with the longest record. Hundreds of thousands of new industrial robots are installed every year, and the world now runs more than four and a half million of them. Automation changes which jobs exist, not just how many, and the hard part has long been people and organisations, not the machines.

Maturity

5/ 5
measured
BasisExtremely mature and continuing: 542,000 industrial robots installed in 2024 alone. Its constraints have been human and organisational, not technical, for decades.SourceIFR, World Robotics 2025 (25 Sep 2025); Acemoglu & Restrepo, Journal of Political Economy 128(6), 2020, and Review of Economic Studies 89(1), 2022; E-296 (Venkataramani et al., JAMA Internal Medicine, 2020)Vintage2020–2025
ReadinessCapability 5Deployment 5Absorption 5Bound byskillsTrendflat

Where it feeds in

5 megatrends in 2 forces

Where you have already met this

In US counties where a car assembly plant closed, opioid overdose deaths were 85% higher five years later than in comparable counties. Not every closure was a robot's doing. But that is what a concentrated loss looks like from the inside - and automation produces concentrated losses.

If this moved

If it deepened further, nothing on the map would follow. Every megatrend it feeds sits at 2 or 3, held there by workforce trust, managerial capability and digital transformation. The machines are not the bottleneck.

What to watch for

Time-to-hire for maintenance technicians and system integrators, which is where installation actually stalls. Collaborative robots as a share of new installations.

Not to be confused with

Artificial intelligence. Automation runs steps someone specified; AI infers them. Also not the megatrend hyper-automation of industrial production, which is what happens when this enabler meets everything else.

What it is

The replacement of human execution of a task by machine execution. The oldest category in this Compendium and the one with the longest evidence base.

How it works

  1. 01Someone writes the sequenceEngineers program the steps and their timing into a controller.
  2. 02Sensors check what arrivedCameras and sensors confirm a part is present and correctly placed.
  3. 03Motors moveServomotors drive the robot's axes along a planned path.
  4. 04The tool does the workA gripper, welding gun or spray nozzle at the end of the arm acts on the part.
  5. 05The result is checkedVision systems verify the part before the line moves on. Every step is fixed in advance - which is what makes it reliable, and what makes it rigid.

State of play

  • 542,000 industrial robots were installed worldwide in 2024, and the operational stock reached about 4.66 million (IFR, World Robotics 2025, 25 Sep 2025).
  • In 2021, 12% of US manufacturing establishments used robots, and 31% in transportation-equipment manufacturing (US Census Bureau data, as reported by EIG, 19 Dec 2025).
  • One more robot per thousand workers reduced the employment-to-population ratio by 0.2 percentage points and wages by 0.42% across US commuting zones, 1990–2007 (Acemoglu & Restrepo, Journal of Political Economy 128(6), 2020).
  • Ageing drives industrial automation, because it creates a shortage of middle-aged workers in manual production tasks (Acemoglu & Restrepo, Review of Economic Studies 89(1), 2022).
  • Five years after a US car assembly plant closed, opioid overdose deaths in the county were 85% higher than in comparable counties (Venkataramani et al., JAMA Internal Medicine 180(2), Feb 2020).
  • US manufacturing had 580,000 job openings in July 2026, a preliminary figure (BLS, Job Openings and Labor Turnover Survey, 1 Sep 2026).

Why it matters

It changes what work exists rather than simply how much. The historical pattern is task displacement and task creation running concurrently at different speeds, and the gap between those speeds is where the social consequences sit.

  • It sits at 5, and every megatrend it feeds sits at 2 or 3. More machines do not steepen the ramp.
  • The losses are local and the gains are national. That is why the politics of automation are worse than the averages suggest.
  • Ageing pushes it forward: where middle-aged workers are scarce, firms automate. Two of the book's forces meet here.

What holds it back

Skills to install and maintain

Integrators, controls engineers and maintenance technicians are scarce, and installation moves at their pace.

Workforce trust

Displaced workers need a credible destination. Without one, resistance is rational.

Managerial capability

Automating a bad process gives you a fast bad process. Redesigning work is a management skill, and it scores 2.

Terms

PLC
A programmable logic controller: the rugged industrial computer that runs a production sequence.
End-effector
The tool at the end of a robot arm - a gripper, welder or nozzle.
Operational technology (OT)
The hardware and software that monitor and control physical equipment, as distinct from office IT.
Degrees of freedom
The number of independent ways a robot arm can move.
Collaborative robot (cobot)
A robot built to work next to people without a safety cage.

Steepens

The gradient behaves like a sticky rubber band: it is pulled hardest toward the weakest enabler. Marginal links are shown but do not pull.

If this moved one step

Scored 5 of 5: there is no step above it.

What if it moved?

What would change the gradient

Conditional, not predictive. No dates, no probabilities.

  1. 01Workforce trust and a credible destination for displaced workers
  2. 02Managerial capability to redesign work rather than substitute within it
  3. 03Relative cost of labour and capital, which policy affects directly

Common misconception

That the debate is about job totals. The measurable effects are concentrated and local - specific occupations, specific places - while the gains are diffuse, which is precisely why the politics are worse than the aggregate numbers suggest.

Talin, B. (2026). Automation. The Future Explorer, dataset v0.4.0 (2026-11-16). https://explorer.talin.digital/e/automation/

Automation: maturity 5/5 · The Future Explorer