By Siddhant Bery, CFA, Head of Strategy, KSP Inc
Imagine a night shift at a mid-sized manufacturing plant. The main press is thumping away, same as it has for twenty years. Somewhere down the line a batch looks wrong. The supervisor checks the clipboard and radios a colleague, and the two spend ten minutes working out where it went off. By then a few hundred parts are in the scrap bin.
That delay is what the factory of 2030 is meant to shrink.
It won’t do it by emptying the floor. The lights-out plant you see on conference slides is not arriving in India any time soon, and plenty of manufacturers wouldn’t want it anyway. The realistic change is smaller. Plants will notice trouble earlier, decide quicker, and spend less time chasing answers by phone.
There is evidence the shift has started. India’s annual industrial robot installations grew by an average of 27 per cent a year between 2020 and 2025, according to the International Federation of Robotics (World Robotics 2026 report). Growth like that rarely stays confined to one or two sectors for long.
Data first, robots second
Ask plant managers what slows them down, and few will name a missing machine. More often it is information scattered across systems that do not agree. Production reports one figure, stores another, and the quality team keeps its own record of rejects.
Artificial intelligence cannot stitch that together on its own. Machines have to be connected, records have to be made consistent, and someone has to be responsible for each data set. IT and operational technology teams, long used to separate budgets and separate meetings, will have to sit at the same table. Manufacturing execution and ERP software can then tie orders to shop-floor activity, purchasing and delivery. The payoff is not another dashboard. It is knowing a batch will miss a customer’s dispatch date while there is still time to act.
Close to the machine, edge computing will do the quick work. A model fitted to a critical motor can pick up a change in vibration or sound and flag maintenance before the breakdown, not after the post-mortem.
Robots, where the sums add up
Robotic welding, handling, inspection and packaging will show up in more plants by 2030, though not evenly. If a job is repetitive and the inputs barely change, automation usually pays. If it means constant changeovers and small batches, skilled hands and flexible kit often still win.
Cobots will probably lead in smaller supplier plants. They sit next to people without heavy guarding, and moving one to a new task takes far less effort. Autonomous mobile robots are taking over the trolley runs that carry parts from racks to assembly stations. One warning for buyers: put a robot in before you’ve sorted out variation upstream, and you will simply produce defective parts faster.
Autonomy with a limit
An autonomous plant does not have to be an unsupervised one. Letting software move a non-critical job when material turns up late is sensible. Changing a safety setting or passing a batch that looks doubtful should stay with a named person.
The stakes rise as plants link machines to enterprise networks, remote service providers and the cloud. A cyber incident in the office is a nuisance. On the production floor it can hurt equipment and people. So keep a list of every connected asset, split the network into segments, and rehearse recovery. Build that in from the start.
The harder part is people
Worries about jobs are fair. Still, what is happening on the floor looks more like a change in the work than its disappearance. The operator who spent shifts making manual adjustments will spend more time reading dashboards and watching how the system performs. Maintenance engineers will have to make sense of sensor alerts. Planners will need the confidence to question what an optimisation tool tells them. IT staff will learn, probably the hard way, that some machines can’t be switched off for an update.
Bigger groups may build platforms across several sites. Smaller suppliers will mostly go machine by machine, using affordable sensors that work with what they already own. Their test is plain: less downtime, fewer rejects, and deliveries that arrive when promised.
The bottom line
The best Indian factories of 2030 won’t necessarily be the ones with the most robots. They’ll be the ones where people trust the data, decisions get made near the problem, and technology helps skilled workers get on with the job. It is less dramatic than the old idea of autonomy and a good deal more useful.