Suyash Sunil
Bhate

Seventeen high-volume lube filter SKUs ran through a single Tier-1 supplier whose cost position had drifted badly out of line. Replacing them was the obvious answer and the frightening one, because these were fast-moving aftermarket parts feeding four distribution centers across five regions. Any gap in supply becomes a stockout, and a stockout in aftermarket is a customer who buys the competitor's filter and never comes back.

I built the transition around the data rather than the calendar. Using SAP MM purchasing info records and MRP demand forecasts, I modelled consumption per SKU and sized the safety stock buffer needed to cover the changeover window at each DC. I rebalanced SKUs across the network so no single site carried the risk, then ran demand-driven replenishment through MMBE and MB52 and tracked every open order in ME2M so nothing slipped quietly. The new supplier was qualified in parallel, not in sequence, so validation never became the critical path.

Whenever sourcing needed to know whether a candidate supplier's air filter actually performed, the sample went out to a lab and the answer came back four days later. That delay was invisible on any dashboard, but it sat on the critical path of every supplier evaluation we ran. Slow validation meant slow onboarding, and slow onboarding meant we kept buying from incumbents by default.

I designed an in-house permeability and ISO-level efficiency test method and brought the capability inside the plant. I designed and 3D-printed the custom test fixtures myself rather than waiting on tooling. To make the results actionable instead of just fast, I built an ALD equivalence tool in MS Access so an engineer could find the right comparable part instantly, and a predictive benchmarking tool in R for cabin air filters.

A production line was quietly losing output, and the default explanation on the floor was equipment capacity. Buying capacity is expensive, slow, and permanent, so before anyone signed a capital request I wanted proof.

I ran a full DMAIC cycle. Measurement and time-motion studies at the station showed the machine was not the constraint at all: cycle times varied widely depending on who was running it. The bottleneck was inconsistent operator method, not horsepower. I deployed SMED to strip waste out of the changeover, standardized the method, then locked the gains into SPC charts and the PFMEA so drift would be visible the moment it started.

Filters were failing in the field for a key customer. Warranty cost was climbing and, worse, the relationship was on the line. Nobody could say why the failures were happening, which meant nobody could promise they would stop.

I led a cross-functional team of four engineers through a structured 8D investigation. Working through Ishikawa and 5 Whys, we isolated the root cause to a reed valve issue driven by assembly torque variation. The fix was not more inspection, it was making the error impossible: I validated a torque-focused Poka-Yoke countermeasure at the station, then updated the standard work, control plans and SOPs before closing the 8D.

Eight workstations carried ergonomic risk: reach distances, working heights and material presentation that were fine for one shift and punishing over a career. They also cost time, because a badly laid out station forces motion that adds nothing to the product.

I redesigned all eight in AutoCAD, working the layouts around reach envelopes and material flow rather than around where the equipment happened to sit. Before committing anything to the floor, I validated the changes in FlexSim simulation so we could see the throughput effect without disrupting live production.

A sugarcane-to-jaggery plant needed a biomass drying system. The existing process took 16 hours per cycle and burned purchased wood and CNG, an ongoing cost on a thin-margin operation. Meanwhile bagasse, the fibrous residue left after crushing cane, was treated as waste.

I ran the full project lifecycle for the HVAC drying system deployment. The core move was closing the loop: engineering dried bagasse into a viable recycled combustion fuel so the plant burned its own byproduct instead of buying energy. I redesigned the process flow around the new heat source and added an automated dry-bulb / wet-bulb moisture monitoring system so drying stopped on measured moisture rather than on a timer.