Aggregate Planning & Scheduling: Strategies, Models & Real-World Trade-offs

Core Concepts: What Is Aggregate Planning?

Aggregate planning is the medium-term (3–18 months) decision-making process that aligns production capacity (workforce, machines, inventory) with demand forecasts to meet customer needs while minimizing costs. Unlike short-term scheduling (which assigns tasks to specific workers/machines), aggregate planning answers:

• How many units should we produce each month?
• How many workers should we hire/fire?
• Should we use overtime, subcontracting, or inventory buffers?

Key Trade-offs in Aggregate Planning

Aggregate planning forces managers to choose between costly but flexible options and cheaper but rigid ones. The classic trade-offs are visualized below:

Why does this matter? In Nepal, Nabil Bank uses aggregate planning to match loan officers to seasonal demand (e.g., more hires before Dasain/Tihar), while Daraz adjusts warehouse staff during sales events like Dashain Sale to avoid delivery delays.

Step-by-Step Aggregate Planning Process

1. Demand Forecasting

Before planning, you need a reliable demand estimate for the planning horizon. Methods include:

• Time-series analysis (moving averages, exponential smoothing)
• Causal models (regression based on economic indicators)
• Qualitative methods (expert judgment, Delphi technique)

Example: NTC’s Workforce Planning NTC forecasts call-center volume using exponential smoothing (weighted average of past data) to predict peak hours (e.g., after Prithivi Jayanti holidays). If demand rises by 20%, they hire temporary agents.

flowchart TD
  A["Step 1: Forecast Demand"] --> B["Time-series\nor Causal Models"]
  B --> C["Adjust for Seasonality\n(Holidays, Festivals)"]
  C --> D["Step 2: Determine Capacity Options"]

2. Capacity Options

Managers choose from five primary strategies to match demand:

Worked Example: Kathmandu Traffic Routes Assume Kathmandu’s Ring Road has:

• Peak demand (7–9 AM): 50,000 vehicles/hour
• Off-peak demand: 10,000 vehicles/hour
• Current capacity: 30,000 vehicles/hour (with 4 lanes)

Aggregate Plan Options:

1. Chase Strategy: Add 2 temporary lanes (cost: NPR 50M) during peak hours.
2. Level Strategy: Keep 4 lanes always (cost: NPR 20M for maintenance).
3. Mixed Strategy: Use dynamic lane control (e.g., reverse lanes at night) + temporary barriers during peak.

Optimal Choice?

• If peak lasts only 2 hours/day → Mixed strategy (lowest total cost).
• If congestion is chronic → Expand permanently (long-term chase).

Mathematical Models in Aggregate Planning

1. Linear Programming (LP) Model

The most common tool for aggregate planning uses LP to minimize costs while meeting demand constraints.

Example: Nabil Bank Loan Processing Nabil Bank wants to plan loan officers for the next 6 months. Data:

• Demand (loans/month): [120, 150, 180, 160, 140, 130]
• Costs:
  • Hiring: NPR 50,000 per officer
  • Firing: NPR 30,000 per officer
  • Overtime: NPR 10,000 per officer/month
  • Inventory (backlogged loans): NPR 2,000 per loan/month

LP Formulation: Minimize: $$\text{Total Cost}=50,000H+30,000F+10,000O+2,000I$$ Subject to:

• Demand constraints: $P_t+I_{t-1}-I_t=D_t$ (for each month $t$)
• Workforce constraints: $W_t=W_{t-1}+H_t-F_t$
• Non-negativity: $H_t,F_t,O_t,I_t\ge 0$

Solution (Simplified):

Total Cost: NPR 2,180,000 (optimal mix of hiring and overtime).

2. Transportation Model

Used when multiple production facilities supply multiple demand points (e.g., Daraz warehouses → cities).

Example: Daraz Delivery Network Daraz has 3 warehouses (Kathmandu, Pokhara, Biratnagar) supplying 4 cities (KTM, PKR, BTN, LTP). Monthly demand and costs:

Objective: Minimize total transportation cost. Solution (Northwest Corner Rule):

• Ship 500 from Kathmandu to KTM (cost: 500 × 5 = 2,500)
• Ship 300 from Pokhara to PKR (cost: 300 × 4 = 1,200)
• Ship 400 from Biratnagar to BTN (cost: 400 × 3 = 1,200)
• Ship remaining 200 from Pokhara to LTP (cost: 200 × 6 = 1,200) Total Cost: NPR 6,100 (per month).

Scheduling: From Aggregate to Short-Term

After aggregate planning, scheduling assigns tasks to specific resources (machines, workers) over weeks/days. Key tools:

1. Gantt Charts

Visualize project timelines (used in construction, IT projects).

Example: NTC 4G Tower Installation

gantt
  title NTC 4G Tower Scheduling (4 Weeks)
  dateFormat  YYYY-MM-DD
  section Site Prep
    Clear Land :a1, 2023-10-01, 3d
    Dig Foundation :after a1, 2, 2023-10-04
  section Tower Assembly
    Assemble Steel :2023-10-07, 5d
    Install Equipment :2023-10-13, 3d
  section Testing
    Signal Test :2023-10-17, 2d
    Final Inspection :2023-10-19, 1d

2. Critical Path Method (CPM) & PERT

Identify the longest path (critical path) to estimate project duration.

Example: NEPSE Stock Trading System Upgrade

graph TD
  A["Start"] --> B["Database Migration\n(5 days)"]
  A --> C["API Integration\n(3 days)"]
  B --> D["Load Testing\n(2 days)"]
  C --> D
  D --> E["Go-Live\n(1 day)"]
  E --> F["End"]

Critical Path: A → B → D → E (11 days total).

In the Real World

1. Nabil Bank: Loan Officer Scheduling

• Idea Used: Aggregate planning (chase strategy) + short-term scheduling.
• How?
  • Medium-term (6 months): Hire temporary loan officers before Dasain (demand spikes by 40%).
  • Short-term (weekly): Schedule officers based on branch footfall data (e.g., more staff on Mondays).
  • Result: Reduced wait times from 2 hours to 30 minutes during peak seasons.

2. Daraz: Warehouse Labor Planning

• Idea Used: Mixed aggregate strategy (inventory + overtime).
• How?
  • Before Dashain Sale:
    • Hire 500 temporary workers (NPR 25M).
    • Increase warehouse shifts to 24/7.
  • During sale:
    • Use just-in-time inventory (no stockpiling).
    • Overtime for 80% of workers (cost: NPR 10M).
  • After sale:
    • Lay off 300 workers, keep 200 on standby.
  • Result: Met 200% demand surge with 15% lower cost than pure chase strategy.

3. NTC: Network Maintenance Crews

• Idea Used: Transportation model for crew allocation.
• How?
  • Problem: 10 repair crews must cover 20 districts with varying fault rates.
  • Solution: LP model assigns crews to districts based on:
    • Distance (cost per km).
    • Fault severity (priority).
  • Outcome: Reduced response time from 48 hours to 6 hours in remote areas.

Common Pitfalls and How to Avoid Them

Exam Tip: How to Score Full Marks

What Examiners Look For

1. Definitions:

   • Clearly distinguish aggregate planning (medium-term) from scheduling (short-term) and capacity planning (long-term).
   • Example:

     "Aggregate planning is a tactical process (3–18 months) that balances production, workforce, and inventory to meet forecasted demand at minimum cost, whereas scheduling is an operational tool (weeks/days) that assigns tasks to specific resources."

2. Mathematical Rigor:

   • For LP problems, always show:
     • Objective function.
     • Constraints (with subscripts $t$ for time periods).
     • Example solution table (like the Nabil Bank case).
   • For transportation models, explain the allocation logic (e.g., Northwest Corner Rule, Vogel’s Approximation).

3. Real-World Application:

   • Link every example to a Nepali company. Examiners reward:
     • Nabil Bank (loan processing).
     • Daraz (warehouse labor).
     • NTC (crew allocation).
     • NEPSE (backlogging trades).
   • Avoid generic examples (e.g., "a car manufacturer"). Be specific.

4. Diagrams and Tables:

   • Draw a trade-off matrix (chase vs. level strategy).
   • Show a Gantt chart or CPM diagram for scheduling questions.
   • Use a transportation cost table for multi-facility problems.

5. Common Exam Questions

   • Short Answer (5 marks):
     • "Differentiate between chase and level production strategies with examples from Nepali industries."
     • "What are the limitations of using linear programming in aggregate planning?"
   • Long Answer (10 marks):
     • "A garment factory in Kathmandu expects demand of [data]. Develop an aggregate plan using a mixed strategy, showing costs for hiring, overtime, and inventory."
     • "Explain how NTC could use the transportation model to optimize its maintenance crew allocation across Nepal’s regions."
   • Case Study (15 marks):
     • "Pathao wants to plan its driver workforce for the next 6 months. Given demand data [table], recommend an aggregate plan using LP, justifying your choice of strategy."

Model Answer Structure for Long Questions

1. Introduction (1 mark):
   • Define aggregate planning and state the objective (e.g., "minimize total cost").
2. Data Summary (1 mark):
   • Restate key numbers (demand, costs) from the question.
3. Strategy Selection (2 marks):
   • Justify why you chose chase, level, or mixed (e.g., "Given high hiring costs, a mixed strategy is optimal").
4. Mathematical Formulation (3 marks):
   • Write the LP objective and constraints.
5. Solution (2 marks):
   • Show a table with $H,F,O,I$ values.
6. Conclusion (1 mark):
   • State total cost and recommend improvements (e.g., "Further reduce costs by negotiating with subcontractors").

Final Visual Summary:

flowchart LR
  A["Aggregate Planning\n(3–18 months)"] --> B["Demand Forecasting"]
  A --> C["Capacity Options\n(Hire/Fire, Overtime, Inventory)"]
  B --> D["Time-Series\nor Causal Models"]
  C --> E["Linear Programming\nor Transportation Model"]
  E --> F["Optimal Plan\n(Cost Minimized)"]
  F --> G["Short-Term\nScheduling\n(Gantt/CPM)"]
  G --> H["Execution\nand Control"]