Who are Toon & Yannick?

The real meaning of dark factories in biotech

"What we often see is that the word 'dark' immediately triggers resistance," Budeners begins. "People think that in Belgium this mainly means cutting jobs and optimizing labor costs. That is certainly part of it, but in practice it is also about quality, reliability, and consistency."

Teyssen observes the same tension with clients. "We often hear companies say they feel they need to do something with dark factories because it appears in strategic roadmaps. But when you ask further, the real question is much more concrete. How do we reduce manual errors? How do we gain more control over our processes?"

To be clear, the lights will not go off in life sciences companies tomorrow. Or the day after. Not out of reluctance, but because biotech is a very different environment from high-throughput industries. Low volumes, high complexity, and strict governance make fully lights-out automation unrealistic. Life sciences companies do aim to reduce human variability. Contamination, inconsistency, and manual errors are the biggest risk factors.

"In practice, we often see automation start with labor cost," says Teyssen. "But automation almost always also focuses on reliability. Systems perform the same task repeatedly in exactly the same way."

This nuance is crucial at a time when the term dark factory appears more often in boardrooms and strategic plans. The hype is real, but the reality is subtler, and that is what makes it interesting.

Automation is more important than ever

Life sciences companies are under pressure to produce smarter, not necessarily bigger. Smarter means better use of data. Every decision, measurement, and deviation must be traceable and defensible.

"We have seen this evolve," Budeners says. "Three or four years ago, some sites could not create reports combining data from two systems. Today, we are talking about central data lakes, real-time sensor data, and dashboards actively used for decision making." For many organizations, this evolution marked the real start of their automation journey.

Because this data foundation is invisible, it is often underestimated. Yet it determines whether advanced automation is feasible. "What we see in practice is that companies look at robots or AI too quickly," Teyssen explains. "Without clear data governance and trust in your data, no system can function reliably."

This highlights a second reality in life sciences: governance is everywhere. A small software change, moving a sensor, or a robot performing a slightly different motion can trigger weeks or months of validation and qualification.

"In a matchstick factory, you make a change and you are good to go," Budeners says. "In pharma, even the smallest changes require a full qualification process. That makes first-time-right extremely important, even though technology benefits from iteration. Clients feel this tension every day."

Why strategy comes before technology

One of the biggest misconceptions about automation is that it is primarily a technological challenge. In reality, the greatest challenges are in culture, adoption, and the ability to start small.

"We often hear: 'We need to do something with robots' or 'with digital twins,'" says Teyssen. "But if you cannot explain which problem you are solving, you are only introducing extra complexity."

Budeners is clear: "Every new technology project creates overhead. Licenses, maintenance, administration, training, all of this is systematically underestimated. Strong projects offset this overhead with cost savings and quality gains. Weak implementations do not."

That is why brownfield sites often offer the biggest opportunities. Not because they are modern, but because they are rich in operational knowledge. "There you see where the real friction is," says Budeners. "Where operators must improvise, where variability arises, and where manual steps introduce risks."

Greenfields may seem more attractive. A blank canvas allows everything to be designed perfectly. But here too there is risk. "In practice, greenfields often include technology that is not sufficiently anchored in real problems," Budeners continues. "In the end, you create more work, not less."

The human factor in dark factories

The question of whether automation makes jobs redundant persists. Budeners and Teyssen often see the opposite. "What we actually see is that automation makes work more intellectual," says Teyssen. "By removing repetitive tasks, operators have space to analyze and make decisions."

Budeners summarizes with a metaphor he often uses: "We need fewer cooks and more chefs. A cook follows a recipe. A chef understands the dish and adjusts."

In the factory, this means less recording and more interpreting. Less executing and more optimizing. Operators become true process owners. This transition is not automatic. Fear of technology, resistance to change, and the tendency to eliminate all risks in advance make projects slow. "That is why we always emphasize involving people from day one," says Teyssen. "Not informing them afterward, but building something together step by step."

How companies can start realistically

The advice Budeners and Teyssen consistently give is simple but demanding. Start small, start with a problem, and start for the right reason.

Starting from strategic questions, such as do we want to produce more with the same infrastructure, reduce variability, work more safely, naturally leads to the right automation projects. These projects are then tested against the reality of the process, infrastructure, and data quality. Automation is not a big bang or a switch. It is a roadmap.

Conclusion

A dark factory in life sciences is not a dystopian vision. It is a mindset that helps reduce human variability, improve data quality, and make companies less vulnerable to labor and cost fluctuations. No lights go out. It is a step-by-step smarter environment in which people are more important than ever.

The future of production is not dark. It is smarter.