Sandeep Girotra, Executive Director and Group CHRO, DCM Shriram in an interaction with Asia Manufacturing Review shared his views on how humans, algorithms and AI agents are working together to redefine the factory floor, what does the rise of agentic AI mean for manufacturing operations, how can manufacturers create collaboration between workers, algorithms and autonomous systems, and more.
Sandeep Girotra is Executive Director and Group CHRO at DCM Shriram, with over three decades of experience in human resources across leading organizations including GE, Ranbaxy Laboratories, HSBC, and DuPont. A postgraduate in Human Resources from Tata Institute of Social Sciences, he has led HR transformation initiatives across India, Japan, Korea, the Middle East, and Singapore.
How are humans, algorithms and AI agents working together to redefine the factory floor?
The factory floor is moving from a machine-led environment to an intelligent work ecosystem. Algorithms are helping teams read patterns from production, quality, maintenance and safety data. AI agents are taking this further by converting signals into action thereby raising alerts, recommending next steps and supporting faster decisions.
But the centre of this shift remains human. People bring context, judgement and accountability. AI can identify a likely breakdown, but a maintenance leader decides the trade-off. AI can flag a quality deviation, but the plant team understands the process reality. The real shift is that employees are moving from reactive firefighting to proactive problem solving.
The question is no longer only “what can be automated?” It is “how should work be redesigned when intelligence is available in real time?” The future factory will be strongest when people, algorithms and AI agents work as one system.
What does the rise of agentic AI mean for manufacturing operations?
Agentic AI marks the shift from dashboards to active decision support. Earlier, systems showed data and humans had to interpret and act. AI agents can now detect issues, recommend actions and trigger workflows within defined boundaries.
In manufacturing, this can improve maintenance planning, quality control, safety monitoring, and production scheduling and workforce deployment. It reduces the gap between insight and action. A possible equipment failure, for example, can be linked to spare availability, maintenance schedules and production impact much faster than before.
However, agentic AI also makes governance critical. Organisations must define what AI can decide, what it can recommend and where human approval is mandatory. Accountability cannot become unclear just because systems become smarter.
The future is not a factory is a human-led, AI-enabled factory where agents bring speed and consistency, while people retain judgement, ethics and responsibility.
How can manufacturers create collaboration between workers, algorithms and autonomous systems?
Collaboration will happen only when technology is built around real work. AI tools must be simple, practical and connected to the decisions workers and supervisors take every day. If AI remains a separate dashboard, it will not change behaviour.
Manufacturers should first identify where AI can add clear value, they must then define the role of each player: what the algorithm recommends, what the system can act on and where people must intervene.
The workforce also must be involved early. Operators and supervisors understand ground realities that data alone may not capture. When they see AI reducing firefighting and improving safety, adoption becomes easier.
The real requirement is trust. Workers should not feel that technology is being imposed on them. They should experience it as a tool that helps them perform better, take quicker decisions and work more safely.
What new skills and roles will emerge as AI agents enter factories?
Manufacturing skills will become more blended. Core technical knowledge will remain essential, but employees will also need digital fluency, data interpretation and the ability to work with AI-led recommendations.
New roles may emerge around predictive maintenance, AI-enabled quality, digital manufacturing coordination, automation support and workforce analytics. Supervisors will also need to evolve. Their role will not only be to manage shifts and output, but to interpret data, act on alerts and coach teams through new ways of working.
For HR, this means skilling cannot remain a training-calendar activity. It has to be linked to how work is changing. Organisations must identify which tasks will be automated, which will be AI-assisted and which human capabilities will become more important.
The most important capability will be learning agility. The factory workforce of the future will need to continuously learn, unlearn and adapt.
Also Read: How Industrial AI Transforms Manufacturing
How are real-time data and predictive analytics improving productivity?
Real-time data is helping manufacturing move from delayed review to early action. Earlier, teams often responded after a breakdown, defect, safety issue or productivity loss had already occurred. Predictive analytics helps identify warning signs before they become serious problems.
This creates practical impact. Maintenance can move from reactive repair to planned intervention. Quality teams can detect process drift earlier. Supervisors can identify productivity gaps during the shift instead of reviewing them later. Leaders can plan manpower, materials and maintenance with better accuracy.
But data by itself does not create productivity. Impact comes when insights are built into daily reviews, escalation routines and decision-making. Dashboards must lead to action.
The next stage of manufacturing productivity will come from foresight - not just knowing what happened, but understanding what may happen next and acting before the loss occurs.
What challenges do manufacturers face while scaling AI?
The biggest challenge is not the AI pilot; it is scaling AI into the operating rhythm of the organisation. A tool may work well on one line or in one plant or a particular function, but scaling it across locations and functions requires data discipline, process maturity and workforce trust.
Data quality is often the first barrier. AI depends on clean, connected and reliable data. If data is fragmented or manually captured, recommendations may not be trusted. Adoption is the second barrier. Workers may resist AI if they feel it does not reflect shop-floor reality or if they see it as a threat.
Governance is equally important. Organisations must define decision rights, escalation rules and accountability. People need to know when to follow an AI recommendation and when to challenge it.
At scale, AI success is not only a technology project. It requires role redesign, capability building, communication, change management and leadership commitment.
How will Industry 5.0 reshape the relationship between people and machines?
Industry 4.0 was largely about automation, connectivity and smart manufacturing. Industry 5.0 brings the human back to the centre. It recognises that factories must be efficient, but also resilient, sustainable and people-centric.
In this model, machines and AI agents will handle more repetitive, data-heavy and precision-led tasks. People will focus more on judgement, creativity, improvement, safety and innovation. The role of humans does not reduce; it becomes more valuable.
For leaders, the priority will be work redesign. Which tasks should machines do? Which decisions should AI support? Where must human judgement remain non-negotiable? These choices will shape the future organisation.
Industry 5.0 also makes reskilling a business priority. Technology should not create a divide between digitally fluent employees and others. It must help people contribute at a higher level.
What will the factory of the future look like?
The factory of the future will be intelligent, connected and human-led. AI agents will monitor operations, flag risks, recommend actions and support faster execution across production, maintenance, quality, safety and workforce planning.
People will continue to own judgement and accountability. Supervisors will spend less time collecting information and more time solving problems. Operators will work with real-time alerts. Engineers will test improvement ideas faster. Quality teams will act before defects become large.
The biggest shift will be shared responsibility. AI may identify the issue, but people will decide priorities, handle exceptions and balance trade-offs between safety, cost, quality and productivity.
The future factory will not be defined by how much technology it has. It will be defined by how well people and technology work together, with AI providing speed, people providing judgement and leadership ensuring trust, learning and accountability.