On-Time Start Rate

Goal of the analysis:

Measure how reliably operations start work orders or operations at their planned time, within a defined tolerance. On-Time Start Rate (OTSR) is a leading indicator of flow stability, schedule discipline, and readiness (materials, tooling, people, and equipment). Executives use it to reduce lead time and variability, improve OTIF, cut expediting/overtime, and decide whether to act on planning rules (frozen windows, sequencing), readiness processes (kitting, maintenance alignment), or asset constraints (SMED, staffing). OTSR is distinct from completion adherence; starting on time without finishing on time signals execution issues, while missing starts signals readiness or schedule realism problems.

Data required:

• Planning and scheduling baselines:
  • Master Production Schedule (MPS), finite capacity schedules, dispatch lists by line/shift, time fences (frozen/slushy/liquid), and start-time tolerance policy (e.g., ±15 minutes).
  • Routings and standard cycle times; changeover matrices and planned setup durations.
• Execution timestamps (MES/ERP/SCADA):
  • Work order/operation planned start vs actual start, planned vs actual finish; produced good quantity and scrap.
  • Predecessor job actual finish, changeover start/finish, first-article release, holds/release events.
• Readiness status and constraints:
  • Material kitting/ATP status (kit complete Y/N; shortages/substitutions), tooling/fixture availability, recipe upload status.
  • Maintenance plan (PM windows), downtime logs (breakdown, starved/blocked), calibration and QA release status.
• Labor and staffing:
  • Timeclock data by crew/shift, skills matrix/certifications, overtime, agency vs FTE, absenteeism.
• Policy and normalization:
  • Exception/override logs (approved reschedules), expedite priorities, business vs clock time, holidays/shift calendars, NPI/engineering change flags.

Detailed step-by-step instruction on how to conduct the analysis:

1. Define metric and scope.
   • On-Time Start Rate (%) = (Number of orders/operations with actual start within ±X minutes of planned start) ÷ (Total scheduled starts) × 100.
   • Track sub-metrics: Early Start %, Late Start %, Median Start Deviation (minutes), Weighted OTSR (weighted by standard hours or demand priority).
   • Define frozen-window horizon (e.g., 24–72 hours) for strict adherence; tag approved reschedules and measure both including and excluding them.
2. Assemble plan vs actual data.
   • Extract last 8–12 weeks of planned and actual starts by line/shift/SKU/operation; align timestamps and time zones; reconcile to production totals.
   • Join readiness data (kit completeness, tooling availability, QA release), predecessor finish and changeover times, and maintenance events.
3. Compute OTSR and timing variance.
   • Classify each start: on-time, early (<−X min), late (>+X min); calculate OTSR, Early %, Late %, and Median/P90 Start Deviation by line/shift and SKU family.
   • For continuous lines, compute “slot start” adherence by takt window.
4. Segment and localize issues.
   • Break down by line/work center, shift/crew, SKU family, changeover pair, and priority tier; rank cells by low OTSR and high business impact (volume/margin/customer criticality).
   • Overlay OEE and plan vs actual output to distinguish readiness vs execution constraints.
5. Attribute root causes.
   • Build a start-loss bridge: Planned start → Setup overrun → Predecessor overrun → Material shortage → Tooling/QA not ready → Labor gap → Maintenance hold → Actual start.
   • Use reason codes and events to allocate minutes to causes; when missing, infer via sequence of events (e.g., kit complete time after planned start).
   • Quantify sequence adherence: % of starts occurring in planned sequence position; high swaps often drive setup overruns and late starts.
6. Link to service and flow.
   • Translate late starts at the constraint into added queue/lead time and OTIF risk; estimate backlog days added.
   • Assess whether early starts inflate WIP or create downstream starvation; flag policy breaches (pulling work early).
7. Trend and stability.
   • Plot daily/weekly OTSR and Start Deviation medians/P90s by line/shift for 12–16 weeks; annotate SMED/TPM events, supplier issues, major expedites or NPIs.
   • Use control charts to detect special causes and prove sustained improvement.
8. Scenario and what-if.
   • Simulate: SMED on top pairs (−30–50% setup), kit completeness to ≥98% (−material delays), frozen window change (24→48h), resequencing by family, added set-up crew or adjusted PM windows.
   • Estimate OTSR lift, lead-time reduction, and OTIF impact; prioritize fastest payback at the constraint.
9. Integrity checks.
   • Ensure “planned start” baseline is the frozen schedule at measurement; exclude orders intentionally moved with approval or tag separately.
   • Validate that starts are not recorded until first value-adding step begins (not staging); cross-check with machine state changes and operator logs.

Format of the output of analysis:

• Executive scorecard: OTSR %, Early %, Late %, Median/P90 Start Deviation (min), Sequence Adherence %, and contribution to OTIF/backlog risk, by line/shift and plant.
• Heatmaps: OTSR by line × shift and by SKU family; red/amber on low adherence cells.
• Gantt variance charts: planned vs actual start times with early/late coloring for top lines/SKUs; sequence swap visualization.
• Loss bridges/Pareto: minutes lost to setup overrun, predecessor overrun, material/tooling/QA not ready, labor gap, maintenance holds.
• Scenario panel: projected OTSR uplift and service impact from SMED, kitting, frozen-window discipline, resequencing, and PM re-slotting.

How to interpret results:

• Low OTSR with high setup overrun minutes: Changeover design and sequencing are primary issues; run SMED and batch by family; enforce realistic planned setup times.
• Low OTSR with many “material/tooling not ready” tags: Readiness process gaps; implement kitting and point-of-use tooling, strengthen ATP and QA release gates.
• Good OTSR but poor completion adherence: Starts are timely, execution is not—focus on reliability (TPM) and centerlining rather than scheduling.
• High Early Start %: Pulling work ahead may increase WIP, changeovers, and downstream starvation; tighten release discipline (CONWIP/FIFO) and frozen windows.
• Shift-to-shift disparities: Skill/capability or leadership variance; standardize work, certify crews, and adjust staffing at changeover peaks.
• Constraint-specific misses: Late starts at the bottleneck disproportionately inflate lead time; prioritize fixes there before non-constraints.

Steps a company can take to improve on this measure:

• Planning and scheduling discipline:
  • Adopt finite-capacity scheduling using demonstrated setup/run times; enforce 24–72 hour frozen windows and clear expedite governance.
  • Sequence by family to minimize “bad neighbor” changeovers; include realistic setup durations in the plan.
• Readiness and material flow:
  • Implement kitting with completeness checks (target ≥98%) and line-side supermarkets; pre-stage tools/change parts on shadow boards; digital start “gates” that require kit/tool/recipe/QA ready.
  • Tighten ATP/CTP logic and QA release timing to avoid last-minute holds.
• Changeover and reliability:
  • SMED to externalize and parallelize internal steps; quick-release mechanisms and presetting; centerline recipes to cut time-to-rate.
  • Align PMs with demand valleys and avoid starting windows; ensure spares and calibration availability.
• People and staffing:
  • Cross-train crews for setup tasks; staff peaks around changeover clusters; cap overtime to prevent fatigue-induced delays; assign changeover champions.
• Governance and digital:
  • Daily tiered meetings reviewing OTSR, start-loss Pareto, and next-24h readiness; real-time dashboards and Andon alerts for impending misses.
  • Lock standards and audit start-time capture; publish “start discipline” KPIs with accountability by line owner.
• Example scenarios:
  • If OTSR is 72% with 35% of misses from setup overruns, run SMED on top 3 pairs and enforce 48h freeze; target OTSR ≥85% within 6–8 weeks.
  • If 40% of late starts cite kit/tool not ready, implement kitting and POUS tooling with digital gates; expect +10–15 pts OTSR and −20–30% lead-time variability.
  • If early starts are 18% creating WIP spikes, deploy CONWIP caps and FIFO enforcement; reduce early starts to <5% and stabilize flow.

Benchmark comparisons:

General benchmarks (directional):

• Frozen-window On-Time Start Rate: world-class 90–95%+; typical 75–90% depending on mix/complexity.
• Median Start Deviation: ≤10 minutes at disciplined sites; P90 within 30–60 minutes for high-mix environments.
• Early Start share: ≤5–10% (policy-driven exceptions only); Sequence adherence ≥85–90% with family batching.

Constructing internal benchmarks:

• Track 12–16 weeks of OTSR and Start Deviation by line/shift/SKU family; publish quartiles and set guardrails (e.g., OTSR <85% for 3 weeks triggers corrective action).
• Measure OTSR with and without approved reschedules; maintain a root-cause registry and close-loop actions; tie improvements to OTIF/backlog metrics.
• Set stage KPIs: kit/QA/tool readiness ≥98% prior to window; setup adherence within ±10–20%; PMs outside start windows ≥95%.
• Rebaseline targets after SMED/TPM, master data updates (standards/changeover matrix), or portfolio shifts; codify best-performing cells as standards.