Supply chain management and ERP

Why it matters

A plant can run a perfect MPS and still miss deliveries because a casting supplier is late or a distributor is sitting on the wrong stock. Supply chain management (SCM) extends planning beyond the factory gate to suppliers, warehouses and customers, and enterprise resource planning (ERP) is the software backbone that carries the same data – orders, inventory, schedules, costs – across all of them. Most graduate engineers in operations roles work inside an ERP system from day one.

Key ideas

Supply chain. All the organisations and activities involved in fulfilling a customer request: raw-material suppliers, component suppliers, manufacturers, distributors, retailers and the customer, linked by three flows:

• Material flow – downstream (and returns upstream).
• Information flow – orders, forecasts, inventory status, both ways.
• Funds flow – payments upstream, credit and returns. SCM is the coordinated design, planning and control of these flows to meet customer service at the lowest total cost (not lowest cost at each stage).

Decisions at three levels.

• Strategic – network design (number and location of plants and warehouses), make-or-buy, supplier base, transport modes.
• Tactical – sourcing contracts, inventory policies, aggregate plans, distribution plans.
• Operational – order promising, scheduling, dispatching shipments.

SCOR model. Processes are grouped as Plan, Source, Make, Deliver, Return (and Enable). It provides standard process definitions and metrics for comparing supply chains.

Push–pull boundary (decoupling point). Upstream of it, stages work to forecast (push); downstream, to actual orders (pull). Make-to-stock places it at finished goods; assemble-to-order at modules; make-to-order at raw material. Postponement moves product differentiation as late as possible.

Bullwhip effect. Variability of orders grows as one moves upstream, even when consumer demand is steady. Causes: demand-forecast updating at each stage, order batching, price fluctuations and promotions (forward buying), rationing and shortage gaming, and long lead times. Remedies: share point-of-sale data, vendor-managed inventory (VMI), collaborative planning, forecasting and replenishment (CPFR), smaller and more frequent orders, everyday low pricing, allocation based on past sales, shorter lead times. It is measured by the bullwhip ratio = variance of orders placed ÷ variance of demand received.

Risk pooling. Holding stock centrally for several markets reduces total safety stock because high and low demands partly cancel. For n identical, independent markets, total safety stock falls by the factor √n (the square-root law), at the cost of longer delivery distances.

Performance measures. Fill rate, order-to-delivery lead time, on-time delivery, inventory turns, days of supply, cash-to-cash cycle time, total supply-chain cost.

Strategic fit. Efficient (low-cost) supply chains suit functional products with stable demand; responsive chains (buffer capacity and stock, fast transport) suit innovative products with uncertain demand.

ERP. An integrated software system with one database shared by all functions: sales and distribution, materials management (purchasing, inventory), production planning (MPS, MRP, CRP, shop-floor control), quality, plant maintenance, finance and costing, human resources. Evolution: MRP (1960s–70s, materials) → closed-loop MRP (capacity feedback) → MRP II (manufacturing resource planning, adds finance and simulation) → ERP (whole enterprise) → extended ERP with SCM, CRM, e-procurement and cloud/analytics.

• Benefits: a single version of the truth, data entered once, real-time visibility, standardised processes, faster closing of accounts, better order promising.
• Risks and issues: high cost and long implementation, need for business-process re-engineering, customisation versus "vanilla" implementation, data migration and accuracy, training and change management, vendor dependence. MRP and ERP output is only as good as BOM, routing and inventory record accuracy.

Formulas

Inventory turns = cost of goods sold per year / average inventory value Days of supply (inventory) = 365 / inventory turns = average inventory / (COGS per day)

• Turns per year; days.

Cash-to-cash cycle = days of inventory + days of receivables − days of payables

• Days.

Bullwhip ratio = Var(orders placed) / Var(demand received)

• Dimensionless; > 1 means amplification.

Pooled safety stock (n identical independent markets) = z · σ · √n versus n · z · σ decentralised

• σ = standard deviation of lead-time demand in one market (units), z = service-level factor (from the normal table).

Worked examples

Example 1 (standard) – inventory turns and cash-to-cash cycle. A pump manufacturer has cost of goods sold ₹120 crore per year and average inventory ₹15 crore. Customers pay in 30 days on average; the firm pays suppliers in 40 days.

1. Inventory turns = 120/15 = 8 per year.
2. Days of inventory = 365/8 = 45.6 days.
3. Cash-to-cash cycle = 45.6 + 30 − 40 = 35.6 days.
4. If lean and VMI raise turns to 12, days of inventory = 30.4 and the cash-to-cash cycle falls to 20.4 days, freeing ₹5 crore of inventory (15 − 120/12).

Example 2 (GATE level) – risk pooling and the bullwhip ratio. (a) A company serves four regions from four warehouses. Weekly demand in each region is independent with σ = 100 units; lead time 1 week; cycle service level 95 % (z = 1.645).

1. Decentralised: SS per warehouse = 1.645 × 100 = 164.5 units; total = 4 × 164.5 = 658 units.
2. Centralised: σ of total demand = 100 × √4 = 200 units; SS = 1.645 × 200 = 329 units.
3. Saving = 1 − 329/658 = 50 % = 1 − 1/√4, as the square-root law predicts. (b) A retailer's weekly customer demand has σ = 20 units, but its weekly orders to the distributor have σ = 50 units.
4. Bullwhip ratio = 50²/20² = 2500/400 = 6.25 – order variance is more than six times demand variance, a strong sign of batching or forecast-updating effects.

Common mistakes

• Thinking that minimising cost at each stage minimises total supply-chain cost.
• Taking the bullwhip ratio as a ratio of standard deviations; it is defined on variances (or state clearly which is used).
• Applying the square-root law to correlated demands; pooling gains shrink as correlation rises.
• Computing inventory turns with sales revenue in the numerator and inventory at cost in the denominator (mixing bases).
• Treating ERP as just software: most failures come from process, data and people issues.
• Assuming JIT and lean remove the need for any buffer; critical, long-lead parts may still need strategic stock.

For GATE PI

Expect conceptual questions on supply-chain flows, the bullwhip effect and its causes and remedies, push–pull systems, VMI and CPFR, the evolution MRP → MRP II → ERP, and ERP modules. Short numericals may cover inventory turns, days of supply, risk-pooling safety stock and bullwhip ratios. Practise stating causes and remedies of the bullwhip effect precisely.

Quick check

1. COGS ₹60 lakh per year, average inventory ₹10 lakh. What are the inventory turns and days of supply?
2. Nine identical independent markets are pooled into one warehouse. By what factor does total safety stock change?
3. Name two causes of the bullwhip effect.
4. What did MRP II add to closed-loop MRP?

Answers: 1. 6 turns per year; about 61 days. 2. It becomes 1/3 (√9 = 3). 3. Any two of: forecast updating, order batching, price promotions, shortage gaming, long lead times. 4. Financial planning and business-wide resource planning (and what-if simulation).