We talk a lot about data and AI — but their real value is human and collective. That's the guiding thread of this panel at Tech For Industry Show 2026 (Paris Expo Porte de Versailles), bringing together Delphine (Forvia), Michaël Valentin (OPEO) and Mathieu Cura (Optimistik) to examine how the 4.0 transformation concretely changes how organizations decide, act and learn — beyond productivity gains alone.

Each speaker sums up their conviction in one sentence. For Forvia, the real hero is the useful time given back to teams to decide better and serve customers better. For OPEO, the key to success in a volatile post-Covid world is to quickly scale local successes — and the good news is that there's no industrial organization where nothing works well somewhere. For Optimistik, once data becomes part of the organization's common language, you decide more accurately and faster.

Delphine identifies four levers. First, the burning platform: start from a concrete, irritating, high-business-impact problem, with a clear use case and KPI. Then standardization and data quality. Then alignment with IT, which brings specialized skills. Finally, the leader: supporting teams daily and ensuring the freed time is useful and grows the teams.

Michaël Valentin insists: you must drive transformation through the business, prioritizing the topics "that keep the leader up at night." Second point: dig into lean's "hidden drawers" — much remains to be done on standardization, flow agility and real-time management. Third, decisive point: integrate digital/AI and operational excellence, because "there's no point digitalizing waste." Those who succeed don't separate a "shiny" Chief Digital Officer on one side and operational excellence on the other: they do both together, transforming the practices of all trades and hierarchical levels.

For Mathieu Cura, setting tools aside, performance rests on three pillars: domain expertise, method (keeping the good practices and fundamentals that work), and a managerial pillar of change management and facilitation. Without movement and action, no ROI — but if these fundamentals are in place, data and AI fluidify, accelerate and multiply the gains. Far from replacing humans, AI strengthens the expert, who feeds it with the right vision and data.

Delphine describes a series of iterations, not a single example. On burning platforms of cost reduction and non-quality (cash impact), Forvia first brought teams on board via a one-day hackathon gathering 50 people from different divisions, cultures and trades (leather, plastics, metal). Quick finding: taxonomy and data architecture were critical. Hence a first step of digitalizing reporting to free up time, then upskilling teams on problem solving (also digitalized), validated on representative pilot factories (seating, plastic injection, door panels, dashboards) before scaling across the 240 factories. Early this year, standard prompts made the data reliable. The result: tens of millions of euros in annual gains (direct cash and avoided costs) on €20 billion in revenue. Next step: linking this operational data to customer verbatims from the customer experience.

Mathieu Cura takes a food-industry client where data changed the job, not just the management. By automating collection (paper forms for process control, quality, regulatory traceability), you meet requirements and reposition the operator as a line pilot: parameters are under automatic control (control charts), only deviations are surfaced, and the operator shifts from a passive posture to an improvement approach. Above all, centralized, contextualized information with a common vocabulary (taxonomy/ontology) "puts the church back in the middle of the village": it breaks down silos between production, maintenance and quality — historically siloed with their own tools — and by de-siloing information, de-siloes the organization. The time freed from "muda" (waste) is reinvested in value-added, revaluing and remotivating tasks.

Michaël Valentin takes a machine builder, the last European player in its niche, ~30% more expensive than Chinese competitors — so ostensibly a cost-reduction issue. Rather than cutting line by line, the client asked a global question: what gives the Chinese their advantage? Hence four priorities: product platformization/modularization (capitalizing on earlier versions), flow agility (accelerating lead times and improving quality, since solving a problem requires understanding what happened within 24 hours), trade excellence (notably maintenance, scarce and slow to train), and reducing support costs (tender responses accelerated by AI). A telling anecdote: a maintenance technician had "cracked" the CMMS at home, in Python, to rebuild a dashboard suited to his needs — proof that every organization has talents frustrated at not having at work the tools they use at home. Overall result: the company is becoming profitable again.

Delphine confirms the field impact: digital problem declarations (scan, automatic defect characterization by camera) make field managers' work easier, surface more problems, reduce line stoppages and the turnover tied to the weariness of "always doing the same thing." Freed time is reinvested, for example in accelerating preventive maintenance, giving some people higher-value tasks — a virtuous, revaluing cycle. As the moderator sums up: being in the approach means being an actor — and when you're an actor, you perform more.

In conclusion, the speakers converge on a broader reading of value. OPEO offers a three-word mindset: stay stoic (accept what you can't control, transform what you can), play collective (connect startups, institutions and manufacturers), and show tenacity (pursue foundational work like data structuring, even when priorities change). Optimistik distinguishes two ROIs: the classic per-use-case ROI (productivity, energy, materials) and a transformation ROI, multiplied by the number of people whose capability you augment — hence the watchword: adoption. Forvia concludes: if the 4.0 transformation doesn't concretely change your decisions and how you work, it won't change your results. The moderator's closing line sums it up: the real ROI of 4.0 is that technology disappears and collective capability appears.

Three exchanges with the audience round out the picture. On freed time, Forvia directs it toward upskilling (Forvia University) and OPEO recalls you need fine enough management to capture it, otherwise it "vanishes"; a Japanese group's example shows you can pull full-time equivalents out of production not to cut jobs, but to prepare the future (digital transformation, new products). On visual inspection, Forvia recalls inspection is itself a waste to reduce, while deploying LLM computer-vision solutions on its "mother factories" then at scale, and cites predictive quality (lab checks cut threefold at one client); OPEO notes maturity has "exploded" over 3 years, the remaining barriers being mostly behavioral (address the root cause, don't stack checks "to reassure"). On LLM limits in quantitative tasks, Forvia combines specialized internal teams and external partners, because "it's moving so fast" that specialized startups remain indispensable.

Why isn't 4.0 transformation only technological? Because its success is above all human and managerial: it requires a clear burning platform, data standardization, alignment with IT and leadership that transforms the practices of all trades. Digitalizing waste adds no value.

What to do with the time freed by automation? Reinvest it in value-added tasks: problem solving, upskilling, preventive maintenance, preparing the company's future. This requires fine management, otherwise the freed time isn't captured.

How does data break down organizational silos? By centralizing and contextualizing information with a common vocabulary (taxonomy/ontology), it gives production, maintenance and quality a shared language — which, by de-siloing information, de-siloes the organization itself.

Will AI replace domain expertise? No, according to the panelists: AI strengthens the expert, who feeds it with the right vision and data. Bad data "multiplies chaos"; data reliability is therefore the condition for real problem solving.

This panel was recorded at Tech For Industry Show 2026, the B2B trade show for industrial transformation (Smart Manufacturing, AI, Supply Chain) at Paris Expo Porte de Versailles. Watch the full discussion above, and join us on May 19–20, 2027, Hall 7, for the next edition.