Mapping the process. Finding the opportunity. Delivering the improvement.
Company names and identifying details have been omitted or generalized under client confidentiality agreements. Metrics and methodology reflect real project work.
Automotive component manufacturer. The press stopped production every time it switched between two parts — a brake bracket and a suspension bracket — costing roughly 9.5 hours of lost production per day across three changeovers. Goal: cut changeover time from 190 minutes to under 60, with no capital investment.
Changeovers averaged 190 minutes, occurring 3 times per day — roughly 9.5 hours of lost production time daily. Target: reduce to under 60 minutes without capital investment.
Filmed the full changeover and split every activity into internal (machine must be stopped — 160 min) vs. external (can happen while running — 30 min). Waste breakdown: 35 min waiting, 18 min walking (~480m per changeover), 12 min searching for parts, 22 min trial-and-error on machine settings.
Fishbone analysis pointed to five root causes: no standard changeover sequence between operators, manual alignment and sensor adjustment, no checklist or setup cart, bolts and tools stored separately from the press, and machine settings that were never documented.
Seven changes: moved die retrieval to external time (−20 min), built a pre-staged SMED tool cart (−15 min), replaced 12 bolts with quick hydraulic clamps (−13 min), added fixed locating blocks instead of manual sensor adjustment (−9 min), introduced a standard cleaning kit (−7 min), moved machine settings to a PLC recipe (−19 min), and split the job across two operators working in parallel (−14 min).
Standardized the new sequence as documented standard work, added visual management (labeled setup cart, checklists), and locked in machine settings as PLC recipes so operators no longer enter them manually.
Final assembly of a precision electro-optical unit (machined housing, optical lens, internal electronics, alignment, sealing, and functional test) was exceeding takt time and producing elevated defects at final inspection, with high and rising rework hours.
Baseline: 95-minute assembly cycle time, 91% first pass yield, 9% defect rate. Goal: cycle time to 60 min, defects below 3%, first pass yield above 97%.
Broke the 95-minute cycle into 8 operations via time study. Defect Pareto across 180 monthly defects (of 2,000 units) showed optical alignment failure and lens contamination as the two largest categories, followed by fastener errors, cosmetic housing damage, seal issues, and electrical connections.
Time studies and workstation observation found ~12 min/unit of motion waste (walking for tools, fasteners, gauges) and ~10 min/unit of waiting (inspection approval, shared tools, alignment equipment). A 5 Whys on alignment failures traced the root cause to manual, experience-based alignment with no standardized fixture — the process depended on individual operator skill rather than a controlled setup.
Six changes: point-of-use kitting per unit (−10 min), 5S workstation redesign with shadow boards and dedicated fixtures (motion waste 12→3 min), standard work instructions with photos and torque specs, a precision alignment fixture with defined reference points (alignment defects 55→10/month), error-proofed fastener kits with torque-controlled tools (fastener defects 31→5/month), and in-process quality checkpoints added after optical and electronics installation instead of only at final inspection.
Weekly standard work and 5S audits, monthly defect Pareto review, scheduled fixture calibration, daily torque tool verification and first-pass yield tracking, and annual operator certification.
350-bed regional hospital, 60,000 annual ED visits (~165 patients/day). Rising patient complaints about wait times, along with a growing rate of patients leaving without being seen (LWBS), were hurting both patient satisfaction and hospital revenue.
Average door-to-doctor time was 78 minutes against a 30-minute target, with a 6.8% LWBS rate and patient satisfaction at 72%. Patient survey comments pointed to slow registration, long unexplained waits, and an overcrowded waiting room.
Baseline month of 5,000 patients broke door-to-doctor time into registration (12 min), wait for triage (20 min), triage (10 min), and wait for physician (36 min). A value stream map showed only 34% of total process time was value-added — the rest was waiting.
Pareto analysis found registration delays, physician unavailability, and bed unavailability accounted for ~73% of total delay. A 5 Whys on physician wait time traced the root cause to fixed staffing schedules that didn't flex with the hospital's 5–9 PM arrival peak.
Seven changes: quick registration capturing only essential info upfront (12→5 min), a dedicated triage nurse during peak hours (20→5 min wait), a Fast Track lane for low-acuity patients, a physician stationed at triage during peak hours (36→12 min), 5S at nursing stations, a real-time electronic bed-management dashboard, and demand-based staffing added only during the 5–9 PM peak.
Daily door-to-doctor and LWBS dashboards, weekly registration and 5S audits, monthly staffing and patient satisfaction reviews, and SPC control charts to flag special-cause variation for investigation.
Final assembly line for an automotive Electronic Control Module (ECM), producing 1,000 units/day. Rising customer complaints traced back to defective units escaping final inspection, pushing quality costs up through rework and scrap.
Defect rate was 5% (1,250 of 25,000 monthly units) against a company target below 1%, with first pass yield at 95%. Customer complaints centered on loose connectors, wrong labels, housing gaps, functional failures, and missing screws.
Inspection data showed loose connectors (420/month) and missing screws (310/month) as the top two defect types — 58.4% of all defects combined. At 6 defect opportunities per unit, this worked out to a DPMO of 8,333, roughly a 3.9 sigma process.
Fishbone analysis pointed to inconsistent operator technique, an overdue torque gun calibration, mixed screw lengths, no visual work standard, and unrecorded torque readings. A 5 Whys on loose connectors traced the root cause to no standardized method for verifying the connector was fully locked — operators judged it by feel.
Seven changes: a poka-yoke connector lock with a visual green indicator (420→70 defects), an automatic screw counter that blocks cycle completion until all screws are installed (310→20), barcode label verification against the production order (185→15), digital torque monitoring with out-of-spec alarms, updated standard work with photos and torque specs, operator retraining with practical certification, and 5S at each station.
Monthly torque tool calibration, weekly layered process audits, first-off inspection every shift, barcode verification every lot, and a daily p-chart tracking the proportion of defective units, with any out-of-control signal triggering a root cause investigation.
Illustrative example, built to show the APQP framework in practice rather than a completed engagement. A supplier was preparing to launch a new sensor bracket for a customer's next model year, on a 9-month timeline to start of production (SOP), with zero tolerance for a late or rejected launch.
Reviewed the customer's design requirements and locked in the program timeline, with the DFMEA and tooling design as the two milestones that would make or break the 9-month schedule.
Running the DFMEA with engineering surfaced a bracket mounting point at risk of fatigue cracking under vibration — a failure mode that hadn't shown up on the initial design review. Caught before tooling was cut, it was resolved with a rib redesign instead of a costly mid-program tooling change.
Built the process flow and PFMEA around the revised design, and drafted the control plan so the critical dimension from the DFMEA finding would be checked at every pilot build, not just at final inspection.
The pilot build's Cpk study showed the critical dimension was capable but running close to the lower spec limit. Adjusted the fixture before running the full validation build, which came back well within spec — and the PPAP package was submitted with that data included, ahead of the customer's deadline.
Tracked defect data through the first 90 days of production against the control plan. No corrective actions were needed — the issue caught in DFMEA never resurfaced in the field.
End-to-end ownership of Define–Measure–Analyze–Improve–Control cycles, from charter to control plan handoff.
Minitab for hypothesis testing, regression, DOE, and control charts.
VSM, SIPOC, and swimlane diagrams used to align cross-functional teams before touching a single metric.
Stakeholder alignment and operator training so improvements hold after the project closes — not just at handoff.
Cost-of-poor-quality and payback analysis to prioritize projects finance actually signs off on.
Coached Green Belts and led project reviews.
The core analytical and statistical tools behind the projects above, grouped by where they're used in a project.
Maps the current and future state of a process, separating value-added from non-value-added steps to expose where time and material get lost.
Frames a process at a high level — Suppliers, Inputs, Process, Outputs, Customers — before diving into detailed mapping.
Traces the physical movement of people or material through a workspace to expose wasted motion and travel distance.
Lays out every step of a process in sequence, making handoffs, decision points, and rework loops visible before anything gets measured.
Ranks defect types or causes by frequency to focus effort on the "vital few" contributors instead of spreading attention evenly across all of them.
A simple structured form for collecting frequency data on the shop floor in real time, before it's charted or analyzed.
Shows how a set of measurements is distributed, making it easy to spot shape, spread, and whether a process is centered on target.
Plots two variables against each other to reveal whether — and how strongly — they're related, before assuming a cause-and-effect link.
Breaks aggregated data apart by shift, machine, operator, or supplier to reveal a pattern a single combined dataset would hide.
Organizes potential causes of a problem into categories — People, Method, Machine, Material, Measurement, Environment — to guide root cause investigation.
Repeatedly asks "why" behind a symptom until the true root cause surfaces, instead of stopping at what's visible on the surface.
Rates potential failure modes on severity, occurrence, and detection — in the design (DFMEA) or the process (PFMEA). Uses the current AIAG-VDA Action Priority (AP) framework, which replaced the older RPN calculation in the 2019 harmonized standard, to flag High/Medium/Low priority risks.
Works backward from a failure event using logic gates (AND/OR) to map every combination of contributing causes — common in safety-critical and aerospace investigations where multiple failures must align to cause an incident.
Groups a large set of ideas or issues — often from team brainstorming — into natural themes, used to organize input before analysis begins.
Cross-references two or more sets of items, such as requirements against process steps, to show where relationships and responsibilities intersect.
Tests multiple process variables at once to find which factors actually drive an outcome, instead of changing one variable at a time.
Uses statistical tests to confirm whether an observed difference is real, or just noise in the data.
Measures how well a process performs against its specification limits, showing whether it can consistently meet requirements.
Confirms the measurement system itself is accurate and consistent before trusting the data it produces.
Tracks a process over time to distinguish normal variation from a real, actionable shift.
Documents what gets measured, how often, and by whom, to make sure an improvement holds after the project closes.
Physically or procedurally prevents a defect from happening in the first place, rather than catching it after the fact.
My Lean Six Sigma journey began with a simple belief: most problems are not caused by people, but by processes that have not been designed, measured, or improved effectively. Throughout my career, I became passionate about understanding workflows, identifying the barriers that prevent success, and helping teams turn everyday challenges into measurable improvements.
Lean Six Sigma gave me the tools and discipline to transform that mindset into a structured approach for creating lasting results.
Currently: Available immediately for full-time roles or contract work
Based in: Superior Twp, Michigan, Washtenaw County
Open to: Full-time roles / consulting / both
Industries: Manufacturing, Automotive, Aerospace, Healthcare, and more
Whether it's a role on your team or a project that needs a Black Belt, I'd like to hear about it.