Why the next leap in the manufacturing industry relies on ground-level insights, not just strategy.
What we've learned from speaking to people working in factories
If you spend ample time talking to people in the manufacturing industry (maintenance engineers, quality managers, production supervisors, operators), you observe a pattern that's almost impossible to ignore. Every time "digital transformation" is mentioned, there's a brief pause. It's not resistance. It's not fatigue. It's just realism.
At Zeltask, we've spent countless hours inside food and beverage plants, walking through production lines, sitting with maintenance teams, and interviewing managers and business owners about their daily challenges. And the gap between industry rhetoric and ground-level reality is always wide. Technology advances at an extraordinary pace, but the lived experience inside factories is far more complex.
This tension isn’t theoretical. It is backed by data.
According to Chandu, in his LinkedIn analysis "The Reality of Digital Transformation: Why Most Fail and the Opportunity for Agency AI", it shows that most organizations have automated less than 15% of technically suitable processes for automation. Only 15% have achieved enterprise-wide implementation. Nearly half remain stuck at the pilot stage.
This contrast between ambition and implementation largely defines the current state of AI and analytics in the manufacturing industry.
Key reasons why digital transformation fails in the manufacturing industry
Chandu's LinkedIn publication isn’t academic research, but it reflects what people who have participated in our interviews and fieldwork describe with striking consistency: digital transformation initiatives typically fail for reasons unrelated to the sophistication of technology.
One common issue is that tools arrive before strategy. A new inspection system is installed in the plant or a maintenance module is implemented, but there's no coherent operational logic linking it to the plant's overarching goals. Operators see a tool; managers see a purchase; but nobody sees a unified direction.
Another barrier is how process variability is misunderstood. We’ve seen production lines where the same task is performed differently in each shift, or where stopgap solutions have become the unofficial norm. Automation and artificial intelligence rely on stability; variability undermines reliability long before the first model is trained.
Adoption is another recurring point of friction. Many initiatives invest in software but not in communication, training, or meaningful integration with existing routines. Workers receive new tools with little explanation. Supervisors are asked to change habits overnight. Tools that could add value become another layer of friction.
Governance further complicates the landscape. Plants oscillate between excessive control, which stifles experimentation, and insufficient control, which produces inconsistencies. Without balanced governance, scaling becomes almost impossible.
Finally, too many programs measure activity instead of results. They track the number of automations implemented or dashboards created, but not whether downtime has decreased, compliance improved, traceability strengthened, or inspections made more reliable.
One last issue is that too many initiatives underestimate the importance of onboarding and support. Tools are introduced without giving teams the time, guidance, or context they need to change their work habits. In our own conversations at Zeltask, we’ve seen how essential this is: plants want clear onboarding, access to practical articles, and a support portal to help them adopt new routines at their own pace. Without that foundation, even the best-designed tools will struggle to take root.
How the manufacturing environment is changing in 2025
Beyond internal challenges, the manufacturing landscape itself is evolving at a pace that would be difficult for any organization to absorb alone. According to Industrial-Production.de, 2025 marks the early but significant emergence of Industry 5.0, a phase defined not only by automation but by deeper collaboration between humans, machines, and environmental systems. AI is not treated as an accessory but as a fundamental capability tying these components together.
There’s a significant shift occurring as manufacturers move from isolated AI tools to integrated architectures. In our own discussions with plant managers and owners, we see this change reflected in the growing interest in unified systems rather than fragmented, department-specific software. People want fewer platforms and more connected information.
Sustainability is also becoming operational, not conceptual. Manufacturers are starting to use AI to identify real intervention points: energy anomalies, waste patterns, bottlenecks; rather than chasing symbolic green initiatives. This aligns closely with what quality and safety teams tell us: "Show me where the real risk is, and I'll act."
Labor shortages add another layer of pressure. A study from The Manufacturer quoted by Columbus Global reports that 97% of manufacturers struggle to hire skilled labor. This reflects what we constantly hear, especially in food production: teams are overburdened, turnover is high, and knowledge disappears with retirements.
Decision-making is also shifting towards integrated analytics. Plants want fewer blind spots. They want incidents, inspections, maintenance records, and machine data to tell a unified story. This is precisely what operational visibility platforms like Zeltask aim to offer.
Why AI pilot projects are growing but not their successful scaling
Across the industry, AI adoption is increasing at a remarkable pace. However, as highlighted by a report featured by TechMonitor, 56% of European manufacturers remain stuck in the pilot phase. Large companies are scaling slowly. Smaller companies are barely scaling.
Yet the momentum is undeniable. Silicon Saxony reports that 42% of German manufacturers are already using AI in production, and another 35% intend to adopt it. In the UK, findings from Rockwell Automation show that 88% of manufacturers have invested or plan to invest in AI within the next year.
The global growth trajectory is even more spectacular. AI in manufacturing is expected to grow from $7.6 billion in 2025 to $62.33 billion in 2032, with Asia-Pacific leading in adoption. European companies are catching up, and use cases are becoming more diverse, from production optimization to marketing and sales.
Most importantly, the results are measurable. McKinsey reports that companies implementing machine learning are three times more likely to improve key performance indicators. PwC notes that 98% of industrial companies expect digital technologies to improve productivity in the coming years.
This reflects what we observe in food and beverage plants: interest is high, potential is high, early results are promising, but scaling requires discipline, process preparation, and unity across all operational areas.
How Europe and Latin America are adopting AI differently
The evolution of AI adoption differs by region.
In Latin America, the transformation is primarily driven by workforce development. A regional report by SAP shows that 81% of large companies in Brazil and Colombia are already investing in AI training. Meanwhile, Nvidia is expanding major AI hubs in Brazil and Mexico, as reported by LatamRepublic, indicating its commitment to building local infrastructure and talent channels. In conversations with Latin American manufacturers, we often hear the same theme: "We need tools, but most importantly, we need people prepared to use them."
Europe, on the other hand, is heavily investing in infrastructure. The European Commission's creation of six new AI factories, backed by 500 million euros (Innovation News Network), demonstrates a strong institutional push towards AI-enabled industry. Advances from Fraunhofer in real-time AI monitoring and predictive quality systems show that applied research is advancing rapidly. However, Accenture's statistics remain a sobering counterweight: more than half stay stuck in the pilot phase.
From Zeltask’s perspective, after talking to clients and partners from both regions, differences are evident. Latin America is building capabilities from the ground up. Europe is creating a top-down environment. Both approaches have their strengths, and both face the same bottleneck when it comes to scalability.
Why digital tools only work when processes and people are ready
All sources, all expert opinions, all conversations within factories lead us to the same conclusion:
Digital transformation becomes reality only when it facilitates daily work, makes it safer, more consistent, or more reliable for the people running the plant.
This belief is deeply embedded in how we approach Zeltask's work. The platform connects inspections, incidents, assets, maintenance tasks, and IIoT information in one place, not because it’s elegant, but because operational teams constantly tell us they're tired of systems that don’t communicate with each other.
They want traceability without paperwork, compliance without chaos, and information without additional steps.
Technology alone doesn’t solve problems. What solves them is understanding the workflow, and that understanding comes from conversations in noisy production rooms and maintenance workshops, not from conferences.
Where AI can generate real operational improvements today
The future of industrial AI won’t be defined by futuristic predictions. It will be defined by the depth with which technology learns to understand the rhythms, constraints, and responsibilities of frontline operations.
The most significant advances emerging from Industry 5.0, from AI copilots to real-time sensor intelligence, revolve around this idea: helping people work smarter, not harder. Supporting technicians during breakdowns. Offering instant visibility to quality teams. Making regulatory compliance seamless. Enabling managers to act with clarity instead of assumptions.
At Zeltask, this is the work we focus on daily: building an intelligent operations platform that helps plants reduce downtime, improve compliance, and bolster continuous improvement by unifying maintenance, quality, and safety into a cohesive system. It's not about digital transformation as a slogan. It's about the daily work of making operations more reliable and more people-centered.
What manufacturers should focus on to successfully adopt digital technology
Manufacturing is entering a new era marked by artificial intelligence, analytics, demographic changes, sustainability pressures, and increasing operational complexity. However, the companies that will stand out won’t be those that merely adopt new tools. They will be the ones that respect the invisible work behind transformation: listening to frontline teams, understanding process variability, investing in adoption, and aligning technology with real workflows.
AI will transform the manufacturing industry. But only if we start where transformation truly occurs: on the production floor.
And that's where Zeltask listens first, before designing, before building, and before implementing anything at all.
Article by
Felipe Borja
CEO & Co-founder
Published on
Nov 15, 2025
