P.1
Gut feel versus data-driven insights
The night-shift quality problem everyone blamed on the crew — until the data showed it was the temperature. Why intuition and evidence work best together, not as rivals.
Preface
Videos
Video hub
Every video from the book, in one place. Watch them here or scan the QR codes in the print edition.
34 videos
Preface
Author’s note
P.2
On overcoming dyslexia with modern tools
How dyslexia, audiobooks and AI tools turned someone who ‘could barely read a book’ into the author of one — the human-machine collaboration story in miniature.
Author’s note
Chapter 2
2.1
Does agile manufacturing make sense?
Can one line really build several different products on the same infrastructure? A look at the convertible factory — and whether the flexibility pays off.
Chapter 2
2.2
Is mechanical engineering becoming a lost art?
AI gets the headlines, but every smart line still stands on precise mechanical and electrical engineering. Why the ‘unsexy’ craft is the real bedrock.
Chapter 2
Chapter 3
3.1
What do the iPhone and Tesla have in common
The feature everyone hyped wasn’t the one that mattered. The hidden architecture lesson behind both the iPhone and the Tesla — and why your line needs the same rethink.
Chapter 3
3.2
Why Industry 4.0 Fails on the Assembly Line (And How to Fix It)
The real reason so many shop-floor digital projects collapse — the Fragmentation Tax — and the architecture that finally fixes it.
Chapter 3
Chapter 4
4.1
No-code line configuration and optimisation
Watch a process engineer configure and change a live line — rules, sequence and all — without writing a single line of PLC code.
Chapter 4
4.2
Building the foundation of the assembly line OS
What it actually means to build the line on a single data model first, and then put the machines on top of it.
Chapter 4
Chapter 5
5.1
How AI Computer Vision is Revolutionizing Manufacturing Efficiency
Teaching the line to see: how AI vision recognises parts and tools in conditions that fixed, rule-based machine vision simply can’t handle.
Chapter 5
5.2
Eliminate mechanical systems with virtual devices
Replacing fixtures, sensors and pick-to-light with a single camera — error-proofing built in software rather than bolted on as hardware.
Chapter 5
5.3
AI virtual devices running on the edge on Nvidia GPUs
Where AI meets the floor physically: vision models running on edge GPUs right at the station, reacting in milliseconds.
Chapter 5
5.4
Teaching the Line to Predict and Prescribe
From forecasting a failure before it happens to recommending the fix — and an honest look at where prediction is genuinely simple and where it’s hard.
Chapter 5
5.5
Demo gone wrong
An honest one: the live AI demo that stalled before it spoke. Why I show the failures, not just the highlight reel.
Chapter 5
5.6
AI Factory Agents That Speak Every Language (Real Manufacturing Use Case)
Meet the factory agents — ask them how production or quality is doing, in any language, all from the same underlying data.
Chapter 5
5.7
Talk to Tom
Asking Tom, the quality agent, how quality is looking — and getting a plain-language answer he worked out from raw data on his own.
Chapter 5
5.8
Many little AIs and one game master (playing noughts and crosses)
Lots of small, narrow AIs and one ‘game master’ that reasons over them all — explained with a simple game of noughts and crosses.
Chapter 5
Chapter 6
6.1
Using technology to enhance job skills and competencies
‘Google Maps for the operator’: how good worker guidance lowers the barrier to skilled work without lowering the quality of the result.
Chapter 6
6.2
Combining technologies to upskill the unemployed
The ladder from no experience to a productive, skilled role — VR, guidance and AI vision combined into a path out of the unemployment queue.
Chapter 6
6.3
Empowering operators with the right data
Data as a signal, not a stick: how the same performance numbers, framed right, turn every operator into a sensor for the whole line.
Chapter 6
6.4
Europe’s Tech Model Fails Africa — Here’s What I’d Do Instead
Why the developed world’s ‘automate people out’ instinct is the wrong aim for most of the planet — and what ‘do more with more’ looks like instead.
Chapter 6
6.5
Automation versus digitisation for different population demographics
Ageing, labour-scarce economies versus young, labour-rich ones — why the same technology should be pointed in opposite directions.
Chapter 6
Chapter 8
8.1
Challenges with Industry 4.0 adoption
The five obstacles that come up again and again — including the one that surprised me most: resistance from the middle, not the floor.
Chapter 8
8.2
LinkedIn poll — why has Industry 4.0 failed to move beyond the pilot stage?
What hundreds of senior people said is really blocking scale — and why I read the popular ‘skills gap’ answer quite differently.
Chapter 8
8.3
How a simple, cheap sensor saved a customer from a costly breakdown
A four-centimetre sensor on a 45-year-old press — proof that the right quick win pays for itself and can breathe life into ageing equipment.
Chapter 8
8.4
Are IIoT sensors for more than just predictive maintenance?
Predictive maintenance is only the start — what else a smart, connected sensor can do once its data flows into the foundation.
Chapter 8
Chapter 9
9.1
Google DeepMind’s Gemini Robots Are Here
Physical AI is arriving: a look at robots that don’t just think but act — and what that begins to mean for the factory floor.
Chapter 9
9.2
Me placing an order with a humanoid (demo at GTC 2026)
From the GTC floor: placing a real order with a humanoid robot — physical AI moving from highlight reels into actual pilots.
Chapter 9
9.3
Humanoids at GTC 2026 (Noble Machines)
Humanoids up close at GTC 2026 — early, but no longer hypothetical, and why they still need a clean data foundation to be useful.
Chapter 9
Appendix A
A.1
VR for operator training
Practising a complex station risk-free in VR — scored, timed, and free to make the mistakes that would be costly on real hardware.
Appendix A
A.2
A cloud-based learning-management system built for VR
Turning VR training from a novelty into a managed programme: assigning simulations, tracking scores, updating content from one place.
Appendix A
A.3
Why 2D Design Reviews Are Failing — And How VR Fixes it
Walking a machine at full scale before it’s built — catching the reach, access and ergonomics problems a flat screen hides.
Appendix A
A.4
Manufacturing wearables — great in theory
The honest reality of AR glasses and smartwatches on the floor — two-hour batteries, change-management surprises, and where they do earn their place.
Appendix A
A.5
Human-machine orchestration
The real productivity isn’t the cobot itself — it’s orchestrating people and machines in sync at a semi-automated station, without compromising safety.
Appendix A
A.6
Has 3D printing revolutionised manufacturing?
A grounded look at additive manufacturing — what it genuinely changes on the floor, and where the hype runs ahead of reality.
Appendix A