When you need to call the parents
Ford has brought back more than 300 experienced engineers / quality specialists after AI quality checks failed to match the expertise of veteran humans engineers.
Ford had leaned into AI across its industrial system, including 900 AI-powered cameras designed to detect quality issues and reduce disruption.
Then came the awkward bit: executives acknowledged that AI trained on design requirements alone lacked the practical judgement of engineers who had lived through multiple product cycles. They made the call to bring back expert to train AI systems, mentor newer staff and hunt failure points before they reached production.
This lands at exactly the moment manufacturers are moving from digital pilots to scaled AI platforms.
KPMG reports that 49% of industrial manufacturing executives already have AI use cases delivering business value and 68% expect to deploy AI at scale within 12 months.
Yet the ILO’s 2026 manufacturing report is the useful cold shower: AI adoption in manufacturing remains uneven, still constrained by skills, data quality, integration challenges and the need for humans to interpret and intervene in complex situations.
The robots and AI don't have enough context or history to manage all situations.
WHY IT MATTERS
This is a story about judgement and the risk of giving control and authority to AI solutions without putting into place the right level of parenting needed whilst the solution learns the context that it needs.
Humans often over-valuing shiny solutions and under-value the tacit expertise that looks 'boring'.
WHAT TO WATCH FOR
The social system around the AI model matters. Look for:
→ Experienced people leave faster than their knowledge is being captured
→ Juniors losing apprenticeship opportunities
→ AI outputs stop being challenged
→ Review forums disappear
→ Leaders cannot answer: “What are my blind spots?”
LIMITATIONS
Much of the public account relies on executive comments and Bloomberg /BBC-linked reporting rather than direct disclosure of performance data.
SOURCE
https://www.bbc.co.uk/news/articles/cgrkd41n2v9o
BESCI AI OPINION
Sometimes you need the engineers that have the tacit knowledge and experience of when things don't go to plan, to save re-learning again and again.
Organisations that downsize often find themselves in an infinite loop of learning as the knowledge leaves the organisation.
Ford got smart, quickly and rehired those with experience.
Will the AI learn and retain - absolutely.
