Transportation Demand, Capacity and Planning

Introduction

Transportation planning converts expected logistics demand into coordinated decisions about vehicles, drivers, infrastructure, schedules, routes, shipment consolidation and transport modes. It is therefore more than the selection of a carrier or the preparation of a dispatch list. It is an integrated planning activity linking sales and operations planning, inventory policy, procurement, warehousing, customer-service commitments and financial control.

A transportation plan must answer five related questions:

1. How much freight will move, where, when and with what service requirements?
2. What transport capacity is available, and how much of it is usable?
3. How should freight be assigned to vehicles, routes and modes?
4. How should variability, seasonality and disruption risk be managed?
5. How can cost, service, emissions and asset utilization be balanced?

Demand forecasts are uncertain, capacity is constrained, and transportation decisions are interdependent. A low-cost plan that misses delivery windows may be commercially unacceptable; a high-service plan with excessive empty running may be financially unsustainable. Effective planning therefore combines quantitative analysis with managerial judgement.

Learning Objectives

After studying this chapter, students should be able to:

• forecast freight demand using appropriate time-series and operational methods;
• distinguish theoretical, practical and effective transportation capacity;
• perform fleet-sizing, load-planning and utilization calculations;
• manage seasonality through calendars, capacity reservations and flexible resources;
• evaluate routing calendars, consolidation and mode-selection alternatives;
• connect transportation requirements with integrated business planning;
• assess technology, governance, sustainability and risk implications; and
• interpret transportation plans critically rather than relying on cost alone.

1. Transportation Demand as a Derived Requirement

Transportation demand is a derived demand: it exists because products, materials, returns or people must be moved between locations. Consequently, freight demand is influenced by production volumes, sourcing decisions, inventory positioning, customer orders, delivery frequency, packaging, product density and network design.

A useful planning description of demand includes:

• origin and destination;
• shipment date or time window;
• quantity in units, weight, volume or handling units;
• product characteristics, such as temperature control or hazardous classification;
• service requirement, such as lead time, appointment or reliability;
• loading and unloading constraints; and
• commercial priority and cost-to-serve category.

Transportation demand should not be measured only in tonnes. A shipment may consume capacity because of weight, cube, pallet positions, floor area, axle limits, loading sequence or hazardous-material segregation. The relevant constraint is often the first one reached.

Demand drivers and planning horizons

Strategic planning, usually covering several years, addresses network location, modal strategy, fleet ownership and long-term contracts. Tactical planning, covering months or weeks, addresses carrier commitments, fleet mix, routing calendars and seasonal capacity. Operational planning, covering days or hours, addresses load building, dispatch sequencing, driver assignment and exception management.

The same demand signal can have different meanings at different horizons. A seasonal peak may justify a long-term regional carrier contract, a short-term leased vehicle, temporary labour, inventory pre-positioning or a combination of these actions.

2. Forecasting Freight Demand

Forecasting unit and data architecture

Forecasting begins with a clear unit of analysis. Examples include pallets per lane per day, vehicleloads per region per week, tonnes by mode per month, or delivery stops by customer segment. Forecasts should be generated at a level where demand patterns are sufficiently stable but decisions remain actionable.

Historical data must be cleansed for cancelled orders, stockouts, unusual promotions, strikes, weather events and changes in the customer base. A shipment record should ideally contain order date, requested date, actual dispatch date, origin, destination, quantity, mode, service class and exception reason. Forecast accuracy deteriorates when these fields are inconsistent.

Basic forecasting methods

The naive forecast uses the most recent observation:

where F_t+1 is the forecast for the next period and A_t is actual demand in period t. It is useful as a benchmark, particularly for intermittent or highly unstable demand.

A moving average of order n is:

The method assumes that recent observations are representative and that no strong trend exists. Increasing n smooths random variation but makes the forecast less responsive.

Exponential smoothing is:

where α is the smoothing constant, 0 < α ≤ 1. A high α reacts rapidly to new information; a low value produces greater smoothing. Demand and forecast must use the same units and time interval.

When a persistent trend exists, Holt’s method may be used:

Here, L_t is the level, T_t is the trend, α and β are smoothing constants, and h is the forecast horizon. The method assumes that trend behaviour is reasonably stable over the forecast horizon.

Seasonal transportation demand can be represented by a baseline multiplied by a seasonal index:

where D_t is demand in period t, B_t is the underlying baseline, and S_t is a seasonal index. For a complete seasonal cycle, indices are commonly normalized so that their average is 1.00. If weekly demand is 1,000 pallets and the relevant seasonal index is 1.25, the seasonal forecast is 1,250 pallets, before considering trend or special events.

Forecast evaluation

Forecast accuracy should be evaluated against the decision purpose. Common measures include mean absolute deviation (MAD):

and mean absolute percentage error (MAPE):

MAD is expressed in demand units. MAPE is a percentage but becomes unstable or undefined when actual demand is zero or very small. For intermittent freight, weighted measures, scaled errors or service-level measures may be more appropriate. Forecast bias should also be monitored:

A consistently positive bias indicates systematic under-forecasting and may cause inadequate capacity reservations.

Forecasts should be presented with ranges or scenarios rather than false precision. A base case, high case and low case can be linked to capacity triggers. Forecast accuracy does not eliminate uncertainty; it makes uncertainty visible and manageable.

3. Capacity, Fleet and Load Planning

Capacity concepts

Theoretical capacity is the maximum physical capability of an asset under ideal conditions. Practical capacity allows for maintenance, legal restrictions, breaks, loading delays and operating constraints. Effective capacity is the portion realistically available after commitments, uncertainty and operational interruptions.

For a vehicle, weight capacity is often:

where C_w is payload capacity, W_GVW is the permitted gross vehicle weight, W_tare is empty vehicle weight, W_equipment is the weight of permanent equipment, and W_reserve is a planning allowance. All terms must use the same weight unit. Legal axle limits may make the allowable payload lower than this calculation.

Usable capacity for a planning period can be estimated as:

where N is the number of available vehicles or trips, C_trip is capacity per trip, U is planned utilization, and A is availability after maintenance and other downtime. This is an approximation and should not be interpreted as a physical law. It assumes comparable trips and a consistent definition of utilization.

Capacity utilization is commonly measured as:

Weight utilization and cube utilization should be calculated separately. A vehicle may be cube-constrained but weight-light, or weight-constrained while significant empty volume remains.

Fleet sizing

A simple fleet-sizing estimate is:

where D is demand per planning period, C is capacity per vehicle per period, r is the target loading factor, A is availability, and the ceiling function rounds upward to a whole vehicle. The assumptions are homogeneous vehicles, comparable routes, stable cycle times and no interaction among vehicles. Real fleet plans must additionally test driver hours, legal limits, maintenance, route length, loading windows and return movements.

Worked example. A regional operation expects 2,400 pallet movements per week. Each truck can complete 8 comparable trips per week and carries 20 pallets. Planned loading factor is 0.85 and fleet availability is 0.90.

Effective weekly capacity per truck is:

8 × 20 × 0.85 × 0.90 = 122.4 pallet positions.

Required trucks are:

2,400 ÷ 122.4 = 19.61.

Therefore, at least 20 trucks are required under the stated assumptions. If demand, route duration or driver availability varies materially, a scenario or simulation analysis is preferable.

Load planning

Load planning seeks to use available weight and space while preserving safety, stability, accessibility and delivery sequence. A practical sequence is:

1. validate shipment dimensions, weight and compatibility;
2. group freight by destination, time window and service class;
3. identify the binding constraint—weight, cube, pallets, axles or time;
4. place heavy and stable freight appropriately and respect axle limits;
5. sequence deliveries so that early stops remain accessible;
6. reserve space for priority or regulated freight; and
7. verify the plan against actual vehicle specifications.

The theoretical load-building problem resembles a multidimensional bin-packing problem and is computationally difficult at scale. Operational systems therefore use heuristics, optimization engines and planner overrides. A mathematically dense load is not necessarily a good load if it causes damage, unsafe unloading or missed appointments.

4. Seasonality, Routing Calendars and Flexible Capacity

Seasonality may arise from holidays, weather, harvest cycles, school calendars, promotions, tourism, fiscal-year purchasing or industry shutdowns. The planning response should begin before the peak, because carrier and driver capacity is often least available when demand is highest.

A routing calendar defines recurring movement patterns: origin-destination lanes, dispatch days, cutoff times, delivery windows, vehicle type, carrier and contingency option. It converts individual shipment decisions into a repeatable operating rhythm. A calendar can also support consolidation by assigning compatible orders to fixed departure days.

Capacity strategies include:

• permanent owned fleet for stable base demand;
• contracted capacity for predictable lanes;
• spot-market or brokerage capacity for variability;
• leased vehicles for medium-term peaks;
• intermodal or rail alternatives for suitable long-haul flows;
• inventory pre-positioning before predictable peaks; and
• customer appointment or delivery-frequency redesign.

A sound policy separates base capacity from surge capacity. Reserving all capacity for the maximum forecast is expensive, while planning only for average demand causes service failures. Trigger rules should specify when additional carriers, overtime, mode shifts or customer communication are activated.

5. Routing, Consolidation and Mode Selection

Routing decisions

Routing decisions include direct versus multi-stop service, stop sequence, delivery windows, vehicle compatibility and return-leg planning. A route plan must consider travel time, distance, congestion, tolls, loading and unloading time, driver regulations, road restrictions and uncertain arrival times.

For a route with multiple stops, the objective is not simply minimum distance. A useful generalized cost is:

The terms may be monetary estimates or model components. The objective and constraints must be defined before optimization; otherwise, a route may minimize distance while violating service requirements.

Consolidation

Consolidation combines shipments to improve vehicle utilization and reduce transport cost per unit. It is particularly effective when orders share compatible origins, destinations and delivery windows. Forms include freight consolidation at an origin, hub-and-spoke cross-docking, pool distribution and milk runs.

Consolidation benefits include fewer dispatches, lower empty mileage, improved asset utilization and reduced emissions. Costs include additional handling, inventory waiting, hub operations, coordination complexity and potential loss of responsiveness. Consolidation should therefore be evaluated against total landed cost and service impact, not line-haul price alone.

Mode selection

Mode choice balances cost, speed, reliability, capacity, network coverage, security, product requirements and environmental impact. Road generally provides flexible door-to-door access; rail can offer efficient high-volume long-haul movement; water is economical for suitable bulk or international flows but slower; air provides speed at high cost; and intermodal combines modes to exploit their respective advantages.

A mode-selection matrix should include:

A mode should be selected for the service requirement, not merely for the lowest quoted rate. Deliberate mode substitution can be valuable: urgent or high-value freight may move by air or expedited road, while predictable replenishment may shift to rail, water or slower consolidated road service.

6. Integrated Planning and Transportation Requirements

Transportation requirements planning (TRP) translates supply-chain demand into time-phased transport requirements. It should be linked to sales and operations planning (S&OP), master production scheduling, material requirements planning, inventory deployment and warehouse labour planning.

The integration sequence is typically:

1. create a consensus demand plan;
2. translate demand into production, procurement and inventory flows;
3. convert flows into shipment quantities, lanes and dates;
4. compare requirements with fleet, carrier, driver, dock and terminal capacity;
5. resolve gaps through timing, mode, sourcing, consolidation or capacity changes; and
6. release an executable transport plan with exception thresholds.

A transportation plan should distinguish requested demand from feasible demand. If the plan assumes more dock appointments, drivers or vehicles than exist, the constraint has merely been hidden. Cross-functional trade-offs are essential: producing in larger batches may reduce production cost but create transport peaks; holding more inventory may improve transport consolidation but increase carrying cost and obsolescence.

A useful planning hierarchy is strategic network design, tactical capacity and calendar planning, and operational dispatch execution. Feedback from execution—late deliveries, tender rejection, detention, empty mileage and forecast error—should update future plans.

7. Technology, Data and Artificial Intelligence

Transportation management systems support order consolidation, carrier tendering, route planning, freight rating, appointment scheduling, tracking, proof of delivery and performance reporting. Telematics provide location, speed, fuel, idling and vehicle-condition data. Electronic documents reduce manual errors and improve auditability.

Optimization and machine-learning tools can forecast lane demand, estimate transit times, predict arrival times, recommend consolidation and identify likely capacity shortages. Their value depends on data quality, explainability and operational adoption. An algorithm trained on biased historical allocation may reproduce inequitable carrier opportunities or systematically favour certain regions. Human planners remain responsible for safety, legal compliance, exceptional cargo and decisions that affect customers, drivers or communities.

Important controls include access management, data validation, version control, cybersecurity, model monitoring and documented override rules. Tracking information can expose commercially sensitive locations and must be governed accordingly.

8. Sustainability, Safety and Governance

Transportation planning has ethical and regulatory dimensions. Compliance may involve vehicle weights, hours of service, dangerous goods, driver welfare, emissions standards, road restrictions, customs documentation and data protection. A plan that depends on illegal driving hours or unsafe loading is not an efficient plan.

Sustainability measures may include fuel consumption, emissions per tonne-kilometre, emissions per shipment, empty-mile percentage, mode share and vehicle utilization. Comparisons require consistent boundaries and assumptions. Reducing kilometres may increase handling or inventory, so life-cycle effects should be assessed.

Procurement governance should prevent overreliance on a single carrier, undisclosed conflicts of interest and unfair tender practices. Carrier scorecards should include safety and labour standards, not just price and on-time performance. Responsible planning also considers congestion, noise, local air quality and the impact of delivery policies on communities.

9. Applied Case: Seasonal Distribution of Consumer Appliances

This is an illustrative case. A consumer-appliance distributor supplies 180 retailers from two regional distribution centres. Normal weekly demand is 8,000 pallet movements, but demand rises to 12,000 pallets for six weeks before a major holiday. Retailers require delivery within two days, and the existing fleet provides 9,000 pallet positions per week at a planned utilization of 85%.

Problem

The existing plan cannot meet peak demand. Management is considering additional owned vehicles, spot-market trucking, rail for selected replenishment flows and greater shipment consolidation. However, appliances are bulky, damage-sensitive and often constrained by cube rather than weight.

Analysis

The peak gap is 3,000 pallet positions per week before considering variability. The existing fleet’s capacity is already planned at 85% utilization, so simply loading it more heavily would reduce resilience and potentially create unsafe or late operations. Historical data shows that 20% of peak orders are predictable replenishment orders with flexible delivery windows, while 15% are urgent promotional orders.

The team identifies three segments:

• predictable replenishment: consolidated twice-weekly road departures;
• standard retail orders: scheduled regional multi-stop routes; and
• urgent promotional orders: reserved expedited road capacity.

Rail is rejected for urgent orders because its transit-time variability and terminal handling do not fit the two-day requirement, but it is considered for replenishment flows into a regional cross-dock.

Approach and solution

The distributor retains sufficient owned capacity for base demand, contracts regional carriers for 2,000 peak pallet positions per week, and reserves 1,000 positions through a broker with a pre-agreed price ceiling. Replenishment orders are consolidated on fixed routing-calendar days. Retailers receive appointment incentives for flexible windows. Safety stock is pre-positioned at the regional cross-dock before the peak.

The transport management system applies shipment compatibility rules, cube limits and delivery windows. A daily control tower monitors tender acceptance, available capacity, late-risk alerts, empty mileage and damage. A trigger activates additional spot capacity if forecast demand exceeds the base-plus-contract plan by more than 5% for two consecutive days.

Outcome

The solution does not eliminate peak cost; instead, it makes cost and service trade-offs explicit. Base fleet utilization remains stable, urgent orders receive protected capacity, and flexible orders are consolidated. The distributor reduces emergency dispatches and gains an alternative mode for suitable replenishment, while retaining road flexibility for time-critical flows.

Learning points

The case demonstrates that capacity planning is a portfolio decision. Forecast segmentation, routing calendars, consolidation and pre-positioning can reduce the amount of expensive surge capacity required. It also shows why cube, delivery windows and product handling characteristics must be considered alongside pallet totals.

Challenges and Limitations

Transportation forecasts remain vulnerable to structural breaks, promotions, weather and geopolitical disruption. Historical data may be incomplete or distorted by stockouts and previously constrained capacity. Utilization measures can encourage unsafe overloading or conceal poor service. Optimization models may produce impractical routes when travel-time estimates, loading durations or legal constraints are inaccurate.

Managers should therefore use scenario planning, sensitivity analysis and manual review. A robust plan includes contingency carriers, alternate modes, realistic buffers and clear escalation rules. The best plan is not the one with the highest theoretical utilization; it is the one that performs acceptably across plausible operating conditions.

Best-Practice Checklist

• Define demand in operationally meaningful units.
• Clean and segment historical data before forecasting.
• Use a simple benchmark before adopting complex models.
• Measure bias as well as accuracy.
• Distinguish weight, cube, pallet, axle and time capacity.
• Separate base, contracted and surge capacity.
• Build routing calendars around compatible flows and service windows.
• Evaluate consolidation using total cost and service effects.
• Include empty movements, accessorial charges and handling in mode comparisons.
• Integrate transport planning with inventory, production and warehouse plans.
• Monitor safety, legality, emissions, data governance and resilience.
• Use technology to support—not replace—professional judgement.

Chapter Summary

Transportation demand is derived from wider supply-chain decisions and must be described by quantity, lane, timing, service and product constraints. Forecasting methods range from naive and moving-average approaches to exponential smoothing and seasonal models; accuracy, bias and uncertainty should all be monitored. Capacity must be distinguished as theoretical, practical and effective, with weight, cube, legal and time constraints assessed separately.

Fleet planning converts forecast demand into vehicles and trips, while load planning balances utilization with safety, sequencing and compatibility. Seasonality is managed through forecasting, routing calendars, base-and-surge capacity, consolidation, flexible modes and inventory positioning. Mode selection should consider total cost, service, reliability, capacity, sustainability and resilience. Integrated planning links S&OP, inventory, production, warehousing and transportation requirements. Finally, technology and AI can improve visibility and decisions, but governance, data quality, regulatory compliance, safety and human accountability remain essential.

Key Terms

Transportation demand: Freight movement required to support supply-chain activity.

Derived demand: Demand created by the need to conduct another economic activity.

Freight forecast: An estimate of future shipment volume, timing and characteristics.

Seasonality: Repeating demand variation associated with calendar or environmental patterns.

Forecast bias: Persistent tendency to over-forecast or under-forecast.

Theoretical capacity: Maximum physical capability under ideal conditions.

Practical capacity: Capacity after normal operating constraints are recognized.

Effective capacity: Realistically usable capacity after commitments and uncertainty.

Payload: Weight of cargo legally and safely carried by a vehicle.

Cube utilization: Proportion of available vehicle volume that is occupied.

Fleet sizing: Determination of the number and type of vehicles required.

Load planning: Assignment and arrangement of freight within available transport capacity.

Routing calendar: Recurring schedule of lanes, departures, stops and delivery windows.

Consolidation: Combining shipments to improve utilization or reduce cost.

Mode selection: Choice among road, rail, water, air, pipeline or intermodal alternatives.

Transportation requirements planning: Time-phased conversion of supply-chain flows into transport requirements.

Spot capacity: Short-term transportation capacity purchased as needed.

Empty mileage: Distance travelled without a revenue or planned load.

Transportation management system: Software supporting transport planning, execution and control.

Control tower: Coordinated visibility and exception-management function for logistics operations.

Review Questions

Short Answer Questions

1. Why is transportation demand described as derived demand?
2. Distinguish theoretical, practical and effective capacity.
3. What is the purpose of a routing calendar?
4. State one limitation of MAPE for freight forecasting.
5. Why should weight and cube utilization be measured separately?

Long Answer Questions

1. Explain how demand forecasting supports fleet and carrier-capacity planning.
2. Discuss the trade-offs between owned, contracted and spot transportation capacity.
3. Explain how shipment consolidation can reduce cost while creating service risks.
4. Compare the principal criteria used in transportation mode selection.
5. Describe how transportation requirements planning should integrate with S&OP and inventory planning.

Analytical / Application Questions

1. A company forecasts 1,800 pallet movements per week. Each vehicle completes 6 trips per week and carries 24 pallets. Planned loading factor is 0.80 and availability is 0.92. Estimate the minimum number of vehicles using the fleet-sizing approximation and explain limitations of the result.
2. A retailer experiences demand that is 40% above average during two annual promotional periods. Design a base-capacity and surge-capacity policy, including forecast triggers, carrier arrangements and consolidation rules.
3. Compare direct road service with consolidated intermodal service for a non-perishable product. Recommend a mode under two different service requirements and identify the data needed to justify the recommendation.

Multiple Choice Questions with Answers

Question 1

Which statement best describes transportation demand?

A. It is independent of production and inventory decisions

B. It is derived from the movement requirements of supply-chain activities

C. It is limited to the number of vehicles owned

D. It is always measured only in tonnes

Correct Answer: B. It is derived from the movement requirements of supply-chain activities

Question 2

A moving-average forecast primarily smooths demand by:

A. Using only the highest historical observation

B. Assigning equal weight to a specified number of recent observations

C. Ignoring all historical observations

D. Multiplying demand by a legal weight limit

Correct Answer: B. Assigning equal weight to a specified number of recent observations

Question 3

If actual demand is 500 units and the forecast is 450 units, the forecast error defined as actual minus forecast is:

A. −50 units

B. 0 units

C. 50 units

D. 950 units

Correct Answer: C. 50 units

Question 4

Which capacity measure recognizes maintenance and normal operating interruptions?

A. Theoretical capacity

B. Practical capacity

C. Advertised capacity

D. Nominal shipment demand

Correct Answer: B. Practical capacity

Question 5

A vehicle completes 5 trips per week, carries 20 pallets per trip, has a loading factor of 0.80 and availability of 0.90. Its approximate usable weekly capacity is:

A. 72 pallets

B. 80 pallets

C. 90 pallets

D. 100 pallets

Correct Answer: C. 90 pallets

Question 6

The main purpose of separating base capacity from surge capacity is to:

A. Plan all resources for maximum possible demand

B. Avoid using forecasts

C. Balance stable demand coverage with flexibility for variability

D. Eliminate the need for carrier contracts

Correct Answer: C. Balance stable demand coverage with flexibility for variability

Question 7

Which condition most directly supports shipment consolidation?

A. Incompatible delivery windows and destinations

B. Common origins, compatible destinations and flexible delivery windows

C. Uncertain product classification and no appointment system

D. A requirement for immediate individual delivery of every order

Correct Answer: B. Common origins, compatible destinations and flexible delivery windows

Question 8

Which factor is least appropriate as the sole basis for mode selection?

A. Total cost

B. Required transit time

C. Reliability and capacity availability

D. Quoted line-haul rate alone

Correct Answer: D. Quoted line-haul rate alone

Question 9

Transportation requirements planning should be integrated most directly with:

A. Only vehicle paint specifications

B. Sales and operations planning, inventory and production plans

C. Unrelated corporate branding activities

D. Historical fuel prices only

Correct Answer: B. Sales and operations planning, inventory and production plans

Question 10

Which is an appropriate governance principle for AI-supported transport planning?

A. Allow the model to override legal requirements

B. Use the model without monitoring data quality

C. Maintain human accountability, validation and documented override rules

D. Hide model assumptions from planners

Correct Answer: C. Maintain human accountability, validation and documented override rules