At the Manufacturing and Supply Chain Summit in Dublin, two talks cut through the usual AI noise.
Maria McCann, Digital Transformation specialist and CEO of Future Focused Solutions, speaking from her experience in digital change and organisational psychology, made a direct point: AI does not change organisations on its own. People change the work, if the operating model gives them room to do it. In the room next door at Dublin Tech Summit, Ken Finnegan of Skillnet Ireland made a similar point from a skills angle:
“AI advantage won’t be defined by who has the best technology; it will be defined by who builds the most capable, adaptable workforce around it.”
For manufacturing and supply chain leaders, that should land close to home. The sector has spent years improving systems, controls, reporting and compliance. AI will add another layer of data and recommendation. The question is whether frontline teams can use it in the work, under pressure, across shifts, sites and changing demand.
Technology on top of old work will not go far
Maria McCann warned against placing AI on top of old structures and expecting a different result. Her point was simple. If people are brought in too late, if culture is ignored, if leadership treats AI as a technology project, the work does not change enough.
That is a practical issue on a production line, in a warehouse, in a depot or in a planning team.
A system may recommend a new schedule. Someone still has to understand the impact on the line. A tool may flag a quality risk. Someone still has to know whether that risk matches what is happening at the machine. A dashboard may show a labour gap. A supervisor still needs to know who can move safely, who needs support and who has done the task before.
AI may improve the signal. People still have to interpret it and act.
Maria also spoke about fear. In many organisations, people are worried about what AI means for their role. Some say it openly. Others stay quiet because they do not feel safe enough to ask. On the frontline, that silence matters. It can show up as workarounds, slow adoption, lack of feedback or resistance that managers only notice late.
If a business wants people to help redesign work around AI, it needs a better conversation with the people doing the work. That includes the sceptics. Maria described sceptics as useful because they see risk early. In frontline industries, those people may be the ones who spot the missing step, the unsafe assumption or the exception that a process map missed.
Skills data is now a workforce planning issue
Nicola O’Neill’s talk (Nicola is MD of Harvest Resources of founder of the Digital Learning Institute) focused on the move from jobs to skills.
She started with a point that supply chain and pharma leaders know well. Skills matrices and skills frameworks are already part of the work, especially in production and compliance. The shift now is that skills can no longer sit only in production records. They need to support workforce strategy across the supply chain.
Nicola named the pressures clearly: automation, demand and supply issues, route-to-market change, skills shortages and employee expectations. She also gave a frontline example. In one organisation changing its production line through automation, around 80% of what current team members did was being displaced.
That does not mean 80% of the people disappear. It means the work changes, and the organisation needs a plan for what those people can do next.
This is where many businesses are still exposed. They know job titles. They know reporting lines. They know who completed mandatory training. They may not know enough about real capability.
- Can this person move from packing to goods-in?
- Can this technician coach others on a recurring fault?
- Can this team leader support a new automation process?
- Who has language skills, customer knowledge or site knowledge that is not recorded anywhere?
- Who could step into a short-term project if demand moved?
Nicola put the core question to the room: do you know what skills you need, and do you know what skills you have?
That question belongs with operations and workforce planning, as much as HR.
Static roles are a poor fit for moving work
Nicola contrasted traditional role-based structures with skills-based models. The old model is familiar: job descriptions, departments, hierarchies, reporting lines and hiring for a fixed role.
The newer model starts with a more dynamic view of capability. People have skills profiles. Teams can be built for short-term needs. Work can move across functions. Organisations can find people internally for a project, a peak, a cover requirement or a process change.
This does not mean every factory, warehouse or depot should abandon structure. Nicola was clear that workforce management is not all-or-nothing. The change can start in one area.
That is useful for frontline leaders. A full skills-based operating model may feel too big. A targeted use case is easier to test.
Start with one production cell, one warehouse process, one shift handover problem or one onboarding pathway. Map the skills needed. Capture the practical knowledge. Assess who can do what. Find the gaps. Use that evidence to plan the next move.
Nicola also warned that skills work is a heavy lift. A full transition can take 12 to 18 months. That is why strategic intent matters. A business should be clear about the reason before it builds a framework. Is the goal redeployment, retention, onboarding, automation readiness, compliance, productivity or succession?
Different goals need different levels of detail.
Frontline knowledge is the missing layer
Both talks pointed to the same gap, though from different directions.
Maria focused on people, culture, trust and work redesign. Nicola focused on skills visibility, mobility and workforce planning. For frontline organisations, the connection is knowledge.
The knowledge that makes work happen is often informal. It sits with experienced operators, supervisors, technicians, planners, drivers and team leads. It is shared in quick demonstrations, shift conversations, WhatsApp messages, side notes and repeated explanations.
That knowledge rarely fits neatly into a policy or a learning module. Yet it is the knowledge that helps a new starter avoid mistakes, helps a supervisor cover absence, helps a team adapt to automation and helps a site run consistently when demand changes.
If AI is going to reshape frontline work, that knowledge has to become easier to find and share.
This is where short-form workplace video can help. A supervisor can record a two-minute explanation of a process change. A technician can show the check that prevents a recurring fault. A planner can explain how to read a new dashboard. A warehouse lead can show the correct handling process for a customer-specific requirement.
When those videos are tagged by site, role, process, skill or risk area, they become part of the operating system of the business. They support training, but they also support redeployment, continuous improvement and day-to-day decision-making.
The practical starting point
Manufacturing and supply chain leaders do not need another broad statement about the future of work. They need a way to make the next change easier to execute.
A useful starting point is to ask five questions:
- What frontline work is changing because of AI or automation?
- Which skills will be needed more often?
- Which current skills will matter less?
- Where is the practical knowledge held by a small number of people?
- How will that knowledge reach the next person who needs it?
The answer will usually point to a narrow, useful first project aka your starting point!
That may be onboarding for a high-turnover role. It may be cross-training for absence cover. It may be capturing process knowledge before automation lands. It may be improving handover between shifts. It may be giving supervisors a shared way to explain new AI-supported decisions.
AI will keep changing the work. The companies that handle that well will be the ones that know what their people can do, give them a clear way to learn what comes next, and make frontline knowledge easier to pass on.
