Latest / The Manufacturing Podcast with Fexingo: Factories, Supply Chains, and Industrial Business / How a French Factory Used Lego to Fix Its Safety Culture
Transcript
- Lucas: Luna, I want to start with a number that stopped me cold. In 2023, French automotive parts supplier Novares recorded 14 lost-time incidents at its plant near Lyon. That's injuries serious enough to miss at least one full shift. Luna: That doesn't sound terrible for a factory with, what, 400 employees? Lucas: Four hundred and twenty, roughly. But industry benchmarks for automotive plants in that region are around 2.0 incidents per 100 full-time equivalents. Novares was at 3.3. Their safety committee had tried the usual levers — more signage, mandatory retraining, a points-based penalty system tied to bonuses. Nothing budged. Luna: So what did they do that actually worked? Lucas: They brought in Lego. Specifically, a facilitated workshop method called Lego Serious Play. Over eighteen months, they cut lost-time incidents by 70 percent, from 14 to 4. And the injury severity index dropped even more sharply. Luna: Wait — they had grown adults on the factory floor building with plastic bricks? Lucas: Not on the production line, no. The sessions happened off the floor, in the training room, in groups of 8 to 12. Operators, team leads, maintenance techs. Each person built a model representing a safety risk they saw but felt unable to raise. Then they narrated it. Lucas: The Lego Serious Play protocol has a strict sequence. The facilitator asks a question — 'Build a model of the most dangerous five minutes of your shift.' Everyone builds in silence for three minutes. Then each person explains their model. No cross-talk until everyone has spoken. Luna: That structure forces participation. You can't hide in the back of the room. Lucas: Exactly. And that's the core insight. In a traditional safety meeting, the loudest voices are usually supervisors, and frontline operators nod along. With the Lego method, the hierarchy is suspended. The model speaks, not the rank. Luna: So what kind of risks came out that the company hadn't captured through their incident reporting system? Lucas: One that got a lot of attention internally: a press operator built a model showing that the safety light curtain on his machine was being bypassed during changeovers because the setup took too long under the standard time. Operators were propping the curtain open with a broom handle. That's a near-miss that never shows up in lagging indicators. Lucas: Another group identified that a specific aisle was kept clear per the layout, but during shift handoffs, two forklifts would pass each other there anyway because the route was faster. No one had ever filed a complaint because 'it was fine.' The Lego model made it visible. Luna: So the act of building gave people permission to name problems they'd normalised. Lucas: That's a perfect way to put it. The Novares safety director told me that in the first six months, the plant identified 43 previously unreported hazards through the workshops. Most were low-severity, but a few were genuine latent risks that could have caused serious injury. Luna: And the 70 percent drop in lost-time incidents — that followed after they fixed those hazards? Lucas: Partially. But the more interesting effect was behavioural. Operators started intervening peer to peer. One of the workshop rules is that after building, the group negotiates a shared model of the ideal safe state. That consensus around 'how we want it to be' creates social accountability. People started policing each other in a way that top-down enforcement never achieved. Luna: You know, Lucas, this reminds me of something I've been thinking about regarding listener-supported content. When a show like ours stays independent — no sponsors dictating what we cover — we get to dig into stories like this that aren't the usual factory-automation hype. We can spend 20 minutes on a Lego workshop in Lyon because it genuinely teaches something about organisational behaviour. Lucas: That's exactly why we keep the podcast ad-free. And the way it stays that way is through listeners who toss a few bucks into the kitty at buy me a coffee dot com slash fexingo. It's not a subscription, not a paywall — just a voluntary cup of coffee gesture that adds up. Luna: And we really mean it when we say every contribution helps us keep the editorial calendar independent. If these episodes have ever made you think differently about a factory problem you were wrestling with, that's the reason. Lucas: Alright — back to Novares. The program cost about 15,000 euros for the facilitator training and Lego kits. The plant's annual workers' comp premium was around 120,000 euros before the intervention. After the incident reduction, it dropped by roughly 20 percent. So the program paid for itself in the first year just on premium savings. Luna: Did Novares try to scale this to other plants? Lucas: They did. By early 2025, they'd rolled it out to six of their twelve European plants. Results varied. The plant in Slovakia saw only a 15 percent reduction in lost-time incidents. The facilitator there noted that the cultural buy-in was slower — workers initially saw it as 'playing games' while they had production targets. Luna: So the context matters enormously. What worked in Lyon didn't automatically transfer. Lucas: Correct. In Lyon, the plant manager participated in the very first workshop as a builder, not as an observer. That sent a signal. In Slovakia, the manager sent his deputy and only came for the closing remarks. The Lego Serious Play method is very sensitive to leadership authenticity. Lucas: There's also a Hawthorne effect angle — Lyon was the pilot plant, so it got more attention from headquarters, more follow-up visits, more data reviews. The novelty factor probably boosted the initial results. Luna: But even accounting for that, a 70 percent drop in serious incidents is hard to dismiss as a flash in the pan. Lucas: Especially when you look at the second-year data. Lyon maintained a lost-time incident rate below 1.0 per 100 employees in 2025. That's better than the automotive industry average in France, which hovers around 2.2. So the gains held. Luna: I want to ask about the cost side — 15,000 euros is not nothing for a small plant. Did they measure any productivity lift? Lucas: They did. Unplanned downtime related to safety incidents dropped by 55 percent. That's hours the line would have stopped for investigation, cleanup, replacement worker assignment. The plant's OEE — overall equipment effectiveness — ticked up by about 3 percentage points. Not huge, but meaningful in a just-in-time supply chain. Lucas: There's also an intangible: the HR team reported that voluntary turnover among operators who attended the workshops was 40 percent lower than among those who didn't, over the same 18-month window. Self-selection bias is possible, but the correlation is striking. Luna: Because people feel heard. The workshop is a mechanism for voice. Lucas: Exactly. And voice is a psychological safety indicator. When Amy Edmondson at Harvard talks about psychological safety in teams, one of the markers is whether people feel they can speak up about errors without fear of reprisal. The Lego model essentially builds that muscle. Luna: So what's the takeaway for a plant manager listening right now who's thinking, 'I don't have 15,000 euros for Lego?' Lucas: You don't need the branded bricks. The method is about structured, equal-participation problem-solving. You could do it with index cards and pens. The key constraints are: everyone builds something, everyone explains, no interruption, and the output is a shared artifact that represents the group's consensus. Lucas: What Novares bought was the facilitation rigor. The Lego Serious Play certification trains facilitators to ask the right probing questions and to enforce the protocol. Without that, it's just a craft session. Luna: So the real innovation isn't the toy — it's the social architecture around it. Lucas: Bingo. And that's the part that generalises. Any factory can try a version of this next shift meeting. Ask everyone to draw the most dangerous spot on the floor. Then go around the table. No one can repeat what someone else said. See what surfaces. Luna: I wonder if we'll see more of these behavioural interventions as labour markets stay tight. If you can't easily replace workers, you have to keep the ones you have healthy and engaged. Lucas: That's exactly the calculation Novares made. They calculated the cost of a single lost-time injury — direct plus indirect — at about 8,000 euros. Fourteen injuries cost them over 110,000. The 15,000 euro Lego program was cheap insurance. But the cultural dividend — that's the part that compounds. Luna: Alright, next episode I want to look at a factory that tried the opposite approach — a zero-tolerance disciplinary system — and whether that can ever work in a unionised shop. Lucas: I've got exactly the case for that. Let's do it.