Critical Spare Parts & Service Parts Forecasting

Learning Outcomes

• After completing this module you should be able to:
  • Describe how to forecast demand for mission-critical spare parts.
  • Describe how firms create a strategic differentiator through superior availability and speed of delivery.
  • Explain the unique challenges involved in forecasting demand of mission-critical parts.

Agenda (SCM 5501)

• Critical Spare Parts overview
• Bill’s Furnace mini-case
• Caterpillar Replacement Parts (mentioned)
• IBM Service (mentioned)
• Ford F-150 Parts deep-dive
• Future of Automobile Parts
• United Hatzalah (rapid-response case)

Bill’s Furnace Case (Service Failure Example)

• Context: Bill K. in Alliston, Ontario (snow-belt, −10 °C winters) with wife + three teenage boys.
• Event timeline:
  • Dec 15: Furnace fails; only heat source = wood-burning fireplace.
  • Technician diagnoses a failed circuit board (mission-critical but “uncommon”).
  • Part not in stock; daily follow-ups; family purchases electric heaters (house ≤ $15\,^{\circ}\text{C}$).
  • Jan 30: Replacement board finally arrives; outside temp $-10\,^{\circ}\text{C}$.
• Take-aways:
  • Lead-time for rare critical parts can create severe customer hardship.
  • Service contracts lose value if inventory strategy is inadequate.
  • Illustrates “availability & speed” as competitive differentiators.

Fundamental Question

• “How many parts am I going to need in inventory?”
  • Balances service level versus holding cost.
  • Especially tricky for critical / low-velocity / long-life items.

Case Study – Ford F-150 Critical Parts

Scenario Data (2021 plan)

• Forecasted sales: $24\,000$ F-150 pickup trucks.
• Production rate: $2\,000$ trucks per month.
• Vehicle life: $10$ years ≈ $250\,000$ km.
• Key components analysed:
  • Brake pads
  • Disc brakes: 4 wheels × 2 pads each = 8 pads per vehicle.
  • Pad useful life: $100\,000$ km.
  • Transmission
  • Warranty: $100\,000$ km or 4 years.
  • Useful life: $250\,000$ km.
  • Failure expectation: $10\%$ of units over life.

Required Data Elements for Forecasting

• Forecasted vehicle production/sales by time bucket (monthly).
• Part–to–vehicle usage rate (BOM relationship).
• Component useful life distribution (mean, variance).
• Warranty terms & expected failure rate.
• Vehicle utilisation statistics (km/year, duty cycles).
• Lead-times for part manufacturing & logistics.
• Cost parameters: production cost curves, holding cost, stock-out penalty.
• Service-level targets (e.g., $\ge 95\%$ fill rate within 24 h).

Transmission Demand Calculation (SCM 5501 example)

• Total Lifetime Requirement:
  $24\,000\ \text{trucks}\times1\ \text{transmission}\times\frac{250\,000\ \text{km life}}{250\,000\ \text{km transmission life}}=24\,000$
• Initial Production Need:
  $24\,000\ \text{trucks}\times1\ \text{transmission}=24\,000$
• Replacement Need (failures):
  $24\,000\times1\times0.10 = 2\,400$
• Total Transmissions to Plan For: $24\,000 + 2\,400 = 26\,400$
  • 24 k assembled into new vehicles.
  • 2.4 k stocked as service parts across the 10-year life.

Brake-Pad Demand Calculation (SCM 5501 example)

• Lifetime Pad Sets per Truck:
  $\frac{250\,000}{100\,000}=2.5$ ≈ 3 full pad cycles (rounded up for safety).
• Total Pads Required:
  $24\,000\ \text{trucks}\times4\ \text{wheels}\times2\ \text{pads}\times3=576\,000$
• Initial Assembly Need:
  $24\,000\times4\times2 = 192\,000$
• Service-Lifecycle Replacements:
  $576\,000-192\,000 = 384\,000$ pads stocked over 10 years.

Ordering & Timing Considerations

• When are parts least expensive to produce?
  • During main vehicle production run (economies of scale, shared tooling).
• Holding-cost trade-offs:
  • Storage $$, obsolescence, insurance, tied-up capital.
• Potential strategy:
  • Produce full initial assembly volume during line-build.
  • Produce service‐life stock in staggered batches, front-loading high-failure‐rate years.
  • Apply Economic Order Quantity (EOQ) & multi-echelon inventory models.

Data Collection & Predictive Maintenance

• Connected vehicles generate continuous telematics ⇒ enables:
  • Deeper customer insights (usage, preferences).
  • Operational excellence (inventory & routing optimisation).
  • Real-time response recommendations (roadside support, part pre-positioning).
  • Predictive maintenance: On-board diagnostics signal impending failure so parts can be delivered just-in-time to the repair bay.
  • New digital services (OTA feature unlocks, pay-per-use functions).
• Access models:
  • Direct (embedded sensors, OEM APIs).
  • Indirect (insurance, dealer CRM, workshop networks).

Forecasting Spare Parts vs. Consumer Packaged Goods (CPG)

• Information Needs:
  • Asset installed base, duty cycle, failure distributions vs. POS sales history.
• Constraints:
  • Long product life, intermittent demand, high service-level promises.
• Product Lifecycle Differences:
  • CPG: short life, steady flow; Parts: long-tail demand for >10 years.

Critical Spare Parts – Unique Inventory Model

• Factors:
  • Warranty coverage window (must guarantee stock).
  • Useful life (dictates failure horizon).
  • Purchase timing (bulk with production vs. deferred runs).
  • Customer lead-time: expectation of same-day/next-day replacement.
• Service parts often require multi-period, intermittent-demand forecasting algorithms (e.g., Croston’s method, bootstrapped SBS).

Risk Prioritisation with FMEA

• Purpose: Flush out every possible failure mode, quantify risk, drive proactive action.
• Procedure:
  1. List all conceivable failure modes for each component.
  2. Score each on $1{-}10$ scales for:
  • Severity (impact on customer safety/operation).
  • Occurrence probability.
  • Detection probability (before customer experiences it).
  1. Compute Risk Priority Number (RPN):
     $\text{RPN}=S\times O\times D$ ((1000) worst-case).
  2. Rank & select high-RPN items for corrective action.
  3. Implement design changes, mistake-proofing, better instructions, etc.
• Implication for Inventory: Components with high RPN may warrant higher safety stock or pre-positioning despite low failure frequency.

Ethical, Practical, & Strategic Implications

• Customer welfare: Heating failure in sub-zero climates becomes a health risk; timely spare-part delivery is an ethical obligation.
• Brand equity: Fast service differentiates brands (Caterpillar’s $\langle 97\%\rangle$ same-day parts fill rate is a market advantage).
• Sustainability: Over-stocking ties up resources; under-stocking leads to premature asset scrapping.
• Digital privacy: Leveraging vehicle data for predictive maintenance raises consent & data-ownership questions.

Key Formulas Recap

• Lifetime quantity requirement:
  $Q<em>{total} = N</em>{units}\times U<em>{per\,unit}\times \frac{L</em>{asset}}{L_{part}}$
• Replacement quantity with failure rate $f$:
  $Q<em>{repl} = N</em>{units}\times U_{per\,unit}\times f$
• Risk Priority Number:
  $\text{RPN}=S\times O\times D$
• EOQ (classic):
  $Q^{*}=\sqrt{\frac{2DS}{H}}$ where $D$ = demand rate, $S$ = setup cost, $H$ = holding cost per unit.

Study Tips

• Practice converting narrative data into quantitative requirements quickly.
• Memorise the relationship between vehicle population, component usage, useful life, and failure rate.
• Familiarise yourself with spare-parts forecasting techniques (e.g., Croston, bootstrapping) and when to apply them.
• Re-work the Ford calculations with altered inputs (e.g., 5% transmission failure, 150k km brake life) to solidify comprehension.
• Use FMEA on a familiar household product to internalise risk ranking mechanics.