Lean Manufacturing Consulting: A Roadmap for Indian Plants

A plant head in Pune once described his lean journey to me like this: “We did 5S in 2019, a Kaizen week in 2021, and a value stream map that now hangs in the conference room. Nothing has really changed on the floor.” He was not unusual. Many Indian plants collect lean tools without building a system that makes them stick.

This post lays out a practical roadmap for lean manufacturing consulting: what happens in each phase, what you should expect to see, and where programmes usually go wrong. Use it to run your own programme or to judge a consultant’s proposal.

Start with a diagnosis, not a training calendar

The weakest engagements begin with a classroom. The strongest begin at the gemba, the place where work happens. In the first two to four weeks, the focus should be on understanding how value actually flows and where it leaks.

A sound diagnosis covers:

• Flow and lead time: how long material takes from receipt to dispatch, and how much of that time is waiting.
• Losses on key machines: breakdowns, changeovers, minor stoppages, speed loss and rejection, measured through OEE (Overall Equipment Effectiveness).
• Inventory and space: raw material, WIP and finished goods, plus the floor area they consume.
• Manpower: idle time, walking, rework and dependence on a few skilled operators.
• Management routines: how problems are raised, who reviews them and how often.

Insist on data from the floor, not only from the ERP. In many MSMEs, downtime is logged loosely, and reported OEE of 75% often turns out closer to 50% once stoppages are timed with a stopwatch. That gap is itself a finding, and a useful one.

Link the programme to a business number

Lean fails when it is described as “implementing 5S” or “becoming lean”. It works when tied to a result the promoter or CEO cares about: a 20% cut in lead time, a reduction in rejection cost of ₹40 lakh a year, or higher output without new capital. These figures are illustrative, but the principle holds. Choose two or three targets, convert them into rupees, and review them monthly.

This is also where Hoshin Kanri, the method of cascading strategic goals into departmental actions, earns its place. Even a simplified version stops improvement from becoming a collection of unrelated projects.

Pilot in one value stream

Resist the urge to launch across the whole factory. Select one product family or line that is important to the customer, has visible pain, and has a supervisor who wants change. A good pilot runs 90 to 120 days.

What the pilot should include

1. Current and future state maps for the chosen value stream, showing where waiting, rework and overproduction sit.

2. Foundation 5S in the pilot area, with defined locations, visual standards and daily audits by the area owner.

3. Changeover reduction using SMED (Single Minute Exchange of Die). Separating internal from external activities often cuts changeover time by 30 to 50% without any investment. Treat this as a typical range, not a promise.

4. Basic autonomous maintenance: operators cleaning, inspecting and lubricating their own machines, which exposes hidden defects early.

5. Daily management: a short stand-up at a visual board covering safety, quality, output and issues.

The pilot’s purpose is to prove the method under your conditions, including shift patterns, contract labour and the realities of power fluctuations or erratic supplier deliveries.

Involve operators, or expect relapse

Indian plants face high attrition among operators and a wide spread of education levels. Consultants who design everything in a conference room produce standards that nobody owns. Better practice is to let operators help write the standard work, use one-point lessons with photographs instead of long documents, and recognise contributions visibly. Supervisors need particular attention. They are the link between management intent and shop-floor behaviour, and they are often promoted for technical skill rather than leadership.

When you evaluate a Lean consultant, ask how they train supervisors and how they plan for operator turnover. A good answer includes multi-skilling, simple visual standards and repeat training, not a one-time workshop.

Scale only after the pilot holds

Once the pilot shows stable results for a couple of months, extend it line by line. At this stage, build internal capability: nominate lean champions, train a small team in problem solving, and introduce a structured Kaizen system with a suggestion process that actually closes the loop. Typical scaling tools include:

Sequence matters. Adding kanban to an unstable process only creates stockouts, so stabilise machines and quality first.

Build the sustain mechanism early

Gains erode when the consultant leaves, and sceptical plant heads know it. Sustaining is designed from the start, not added at the end. Look for:

• Layered process audits, where supervisors, managers and the plant head each check at set frequencies.
• Gemba walks by senior leaders with a fixed route and questions, not casual strolls.
• A monthly review of lean KPIs alongside production and finance reviews.
• A formal handover in which your team, not the consultant, leads the meetings.

A TPM or lean audit at six and twelve months is a simple way to catch drift before it becomes the norm.

Choosing the right support

Ask prospective consultants for the practical details. Who will be on the floor, and for how many days a month? Have they worked in your sector, whether auto components, packaging, chemicals or food? How do they measure success, and will they tie fees partly to results? Be wary of anyone promising dramatic gains before seeing your plant. A credible Lean consultant will spend time on the shop floor first and will be comfortable telling you which losses matter most and which can wait.

Conclusion: a roadmap you can start this quarter

The path is straightforward, though not easy: diagnose with real data, tie goals to rupees, pilot in one value stream, involve operators, scale in sequence and design the sustain system from day one. Skipping steps is how plants end up with a well-framed value stream map and an unchanged floor.