Resource planning starts before the work order

This piece follows from production consumes asset life. If production and lifecycle forecasts create future maintenance demand, they also create future resource demand — well before any work order exists.

Maintenance systems often plan resources after work has been identified.

A work order is raised.

Labour is estimated.

Parts are reserved.

A contractor is assigned.

A workshop bay, crane or specialist tool is booked.

That process is necessary.

But by the time the work order exists, much of the resource demand may already be predictable.

Component lives are approaching.

Production forecasts imply higher utilisation.

Condition trends are deteriorating.

Shutdown scope is developing.

Inspection programs are known.

Seasonal work is approaching.

That suggests a useful proposition:

Resource planning should begin with expected future asset demand, not only with work orders that have already been created.

A work order is often a late signal

A work order is an execution record.

It tells the organisation that a specific piece of work now needs to be planned or performed.

But the factors creating that work may have been visible much earlier.

For example:

Production Forecast
      ↓
Expected Operating Hours
      ↓
Component Life Consumption
      ↓
Predicted Intervention
      ↓
Work Order

If an engine rebuild is likely nine months from now, waiting until the work order is generated before considering labour, parts and workshop capacity wastes valuable planning time.

The same applies to:

• shutdown packages;
• statutory inspections;
• road treatment programs;
• structural interventions;
• contractor mobilisation;
• long-lead components.

The earlier the requirement becomes visible, the more options the organisation has.

Resource demand is an output of asset behaviour

Resources are not required because a scheduler decides to allocate them.

They are required because assets create work.

That work may emerge from:

Usage
Condition
Failure
Inspection
Lifecycle
Production
Compliance
Shutdown Scope

Those drivers create:

Expected Work
      ↓
Expected Resource Demand

So resource planning can begin before the detailed job exists.

A useful planning chain is:

Asset Forecast
      ↓
Maintenance Forecast
      ↓
Resource Forecast
      ↓
Capacity Comparison
      ↓
Decision

The work order eventually turns the forecast into an executable transaction.

Labour demand can be forecast

Consider a fleet with known component-life patterns.

The system forecasts the following interventions over the next twelve months:

6 engine rebuilds
4 transmission replacements
18 major services
12 suspension repairs

If historical work standards are known, this can become a labour forecast.

For example:

Engine rebuilds          3,600 h
Transmissions            1,600 h
Major services             900 h
Suspension work             720 h
Other planned work        4,500 h
Forecast reactive work    2,800 h
--------------------------------
Expected labour demand   14,120 h

The exact work orders may not exist yet.

The workload is already becoming visible.

Capacity is not the same as headcount

Suppose a workshop employs twenty fitters.

That does not mean it has twenty fitters available for maintenance work throughout the year.

Actual capacity may be reduced by:

• leave;
• training;
• meetings;
• supervision;
• travel;
• inductions;
• safety activities;
• non-productive time;
• competing work.

So:

Nominal Labour Capacity
      ↓
Availability Adjustment
      ↓
Effective Capacity

For example:

20 fitters
× 1,800 nominal hours
= 36,000 h

Less leave/training/etc.
= 30,500 effective hours

That number can then be compared with forecast demand.

Skill matters as much as total labour

An organisation may have enough total hours and still have a resource problem.

Suppose:

Total labour demand        28,000 h
Total labour capacity      31,000 h

At first glance there is no constraint.

But skill demand may be:

Mechanical fitters         18,000 h
Available                  20,000 h

Auto electricians           6,500 h
Available                   4,800 h

Boilermakers                3,500 h
Available                   6,200 h

The overall labour position is positive.

The electrical position is not.

So resource forecasting needs to understand:

Role
Skill
Qualification
Competency
Availability

not just labour hours.

Qualifications and access can become resource constraints

Industrial work often requires more than nominal trade skills.

A person may also need:

• site induction;
• high-risk work licence;
• confined-space competency;
• working-at-heights qualification;
• electrical authorisation;
• equipment-specific competency;
• medical clearance;
• permit authority.

That means:

Available person

and:

Available qualified person

are not necessarily the same thing.

For a shutdown, this becomes particularly important.

The contractor may supply twenty people.

Only fourteen may currently have all required access and competencies.

The contract is ready.

The resource is not.

Workshop capacity is another resource

Labour is often only one constraint.

A mobile-equipment workshop may also depend on:

Workshop Bays
Wash Bay
Crane Capacity
Component Area
Tooling
Test Equipment
Service Trucks

Suppose the labour forecast is manageable.

But three major component interventions fall into the same month and require the same heavy workshop bay.

The limiting resource becomes:

Heavy Bay Capacity

not labour.

The planning model should therefore be able to identify whichever resource is constraining execution.

Tools and access resources can become critical

Shutdowns make this particularly obvious.

A work package may require:

40 mechanical labour hours

but also:

80 t crane
special lifting beam
scaffold
confined-space access
vendor technician

If one critical resource is unavailable, the labour capacity becomes irrelevant.

This suggests that resource planning should distinguish between:

Consumable Capacity

and:

Critical Enabling Resources

A resource with only one available unit can dominate the schedule.

Parts are also resources

Inventory and procurement are often treated separately from labour planning.

Operationally, they are part of the same readiness problem.

A predicted intervention may require:

Labour
+
Part
+
Tool
+
Access
+
Time Window

If the part has a twelve-month lead time, it may need to be ordered long before the work order is created.

So the planning chain should include:

Predicted Intervention
      ↓
Material Requirement
      ↓
Lead-Time Check
      ↓
Procurement Action

This moves purchasing upstream.

Long-lead items expose the weakness of late planning

Suppose an engine exchange is forecast for March next year.

The supplier lead time is ten months.

If the organisation waits until January to raise the work order, there is already a problem.

The system should be able to recognise:

Forecast Intervention Date
-
Procurement Lead Time
=
Latest Order Date

That turns lifecycle forecasting into procurement planning.

The work order may still be created later.

The purchasing decision cannot wait.

Production plans can create resource forecasts

Production planning is one of the strongest upstream signals.

Suppose production is increased by 15%.

Expected utilisation increases.

That may bring forward:

• component rebuilds;
• services;
• inspections;
• tyre changes;
• planned shutdown work.

The resulting chain becomes:

Production Increase
      ↓
Asset Utilisation Increase
      ↓
Maintenance Demand Increase
      ↓
Resource Demand Increase

Resource planning should therefore respond when production plans change.

That could mean:

• more labour;
• more contractors;
• additional workshop shifts;
• more parts;
• earlier procurement;
• more shutdown scope.

The maintenance organisation should not first discover the consequence when the work orders appear.

Resource planning should also run in reverse

The model can work the other way.

Suppose the organisation has:

Workshop capacity      Fixed
Available fitters      Fixed
Exchange engines       Limited
Contractor availability Limited

The system can ask:

What maintenance demand can we actually support?

That creates:

Resource Capacity
      ↓
Maintenance Capacity
      ↓
Expected Asset Availability
      ↓
Achievable Production

This is important because production plans often assume maintenance capability rather than testing it.

A mature operating model should expose the constraint.

Contractors should be part of the forecast

Contractors are often used reactively.

Internal capacity is exceeded.

A contractor is then brought in.

A better model can predict the shortfall earlier.

For example:

Q2 mechanical demand        10,200 h
Internal capacity            8,500 h
Expected deficit             1,700 h

The organisation can then decide whether to:

Hire
Contract
Outsource
Reschedule
Reduce scope
Change production

The earlier the forecast appears, the better the commercial options are likely to be.

Outsourcing can be treated as a capacity decision

Not all work needs to be performed internally.

For major components, the options may include:

Internal rebuild
External rebuild
Exchange component
New component
Defer intervention

Each consumes different resources.

For example:

Internal Rebuild

Labour      High
Bay         High
Duration    Long
Cash        Moderate

versus:

Exchange Component

Labour      Lower
Bay         Lower
Duration    Shorter
Cash        Higher

The optimum choice may depend on the current resource position rather than simply unit repair cost.

That is a resource-optimisation decision.

Geography changes resource planning

For roads and distributed infrastructure, resource planning has another dimension.

Travel and mobilisation matter.

The requirement might be:

Grader
Roller
Water Cart
Crew
Material

But the sequence should also consider:

Location
Travel
Mobilisation
Weather

The planning question becomes:

Which work should be bundled while these resources are already in the area?

That means:

Work Forecast
      +
Geography
      ↓
Resource Program

This differs considerably from workshop scheduling.

Again, the resource model can be common while the optimisation logic is domain-specific.

Shutdowns concentrate demand

A shutdown creates an extreme form of resource compression.

Work that might normally be spread across months is concentrated into a small operating window.

The shutdown may require:

Mechanical Trades
Electrical Trades
Boilermakers
Scaffolders
Crane Crews
Engineers
Inspectors
Vendor Specialists
Commissioning Teams

The question is not simply whether enough total labour exists.

It is:

Are the right resources available at the right time, in the right sequence, with the right access and supporting materials?

That requires time-phased resource forecasting.

For example:

Day 1
Mechanical demand       220 h
Electrical demand        80 h

Day 2
Mechanical demand       310 h
Electrical demand       160 h

Day 3
Mechanical demand       140 h
Electrical demand       240 h

The overall totals may look manageable.

The peak periods may not be.

Resource optimisation needs time buckets

Annual resource totals can hide short-term constraints.

Suppose:

Annual fitter demand     28,000 h
Annual capacity          30,000 h

The annual position appears healthy.

But monthly demand may look like:

Jan    2,000 h
Feb    2,200 h
Mar    2,400 h
Apr    4,500 h
May    4,200 h
Jun    3,900 h

The constraint occurs in April to June.

So resource positions should be time-based.

Useful horizons may include:

Today
Week
Month
Quarter
Year
3–5 Years

Different decisions belong at different horizons.

Lifecycle planning supports long-range workforce planning

Some resource requirements can be predicted years ahead.

Suppose the fleet lifecycle model shows a major wave of engine and transmission rebuilds in FY29.

That may justify decisions now about:

• apprenticeship numbers;
• training;
• workshop expansion;
• contractor agreements;
• rebuild facilities;
• exchange-component strategy.

This is where lifecycle management becomes workforce planning.

The planning chain becomes:

Long-Term Asset Forecast
      ↓
Expected Work Profile
      ↓
Skill Demand
      ↓
Workforce Strategy

That is much more strategic than weekly scheduling.

Budgeting should follow the resource model

Resource forecasts also create better budgets.

Suppose the organisation forecasts:

Internal fitter hours      30,000 h
Contractor fitter hours     4,000 h
Auto electrician deficit    1,200 h
Crane hire                    900 h

The cost forecast can then derive from expected delivery methods.

The budget becomes:

Expected Work
      ↓
Resource Mix
      ↓
Expected Cost

rather than:

Last Year's Labour Cost
+
Inflation

This creates stronger links between operational and financial planning.

Actual performance should improve future resource assumptions

Resource standards should learn from actual work.

Suppose a particular component replacement is planned at:

80 fitter hours

but recent actuals are:

92 h
88 h
95 h
90 h

The future standard may need revision.

Likewise, if contractor mobilisation repeatedly takes longer than assumed, future readiness models should change.

The feedback loop becomes:

Resource Estimate
      ↓
Actual Consumption
      ↓
Variance
      ↓
Updated Standard
      ↓
Future Forecast

That is consistent with the broader idea that completed work should improve the operating model.

Uncertainty should be visible

Future resource demand is never completely certain.

A useful forecast may therefore distinguish:

Known Work
Probable Work
Risk Work
Contingency

For example:

Next Quarter Labour Demand

Known              7,200 h
Probable            1,900 h
Risk allowance      1,100 h
---------------------------
Expected range     9,100–10,200 h

That may be more useful than pretending the forecast is exact.

Confidence can improve as the work approaches.

Resource planning may need an operational ledger

The operational-ledger idea discussed earlier fits particularly well here.

A resource position might change through entries such as:

New shutdown scope          +600 h
Component deferred          -300 h
Production increase         +450 h
Contractor confirmed      +1,000 h capacity
Leave approved              -120 h capacity

The current resource position becomes explainable.

For example:

MECHANICAL LABOUR – Q3

Internal capacity        8,400 h
Contract capacity        1,500 h
Total capacity           9,900 h

Forecast demand         10,600 h

Expected deficit           700 h

The system can show exactly what created the deficit.

The objective should be decisions, not just forecasts

A forecast without a decision process has limited value.

If a resource constraint is predicted, the system should help evaluate responses.

For example:

Constraint:
Mechanical labour deficit – 1,200 h

Possible actions:

Add contractor labour
Reschedule non-critical work
Outsource rebuild
Move shutdown scope
Add shift
Change production plan

Each option can have consequences for:

• cost;
• risk;
• availability;
• production;
• future workload.

Resource planning therefore becomes a decision-support function.

Different asset domains still need different optimisation logic

The underlying concept is common:

Future Work
      ↓
Resource Demand
      ↓
Capacity
      ↓
Constraint
      ↓
Decision

But the details vary.

Mobile plant

Optimise:

workshop
trades
components
service windows

Roads

Optimise:

crew
plant
materials
geography
mobilisation

Structures

Optimise:

inspection specialists
engineering
access
contractors
traffic management

Processing plant

Optimise:

shutdown crews
critical tools
scaffolding
cranes
isolation
commissioning

The common kernel remains.

The domain-specific model provides the context.

What this could mean for CPM and DSLCore

CPM could treat resource demand as a forecastable operational position.

Useful models might include:

Resource
ResourceType
Skill
Qualification
Capacity
ResourceRequirement
ResourceForecast
ResourceAllocation
ResourceConstraint
ContractorCapacity
ToolRequirement
FacilityCapacity

Then maintenance and lifecycle models can generate future requirements.

For example:

Component Forecast
      ↓
Maintenance Requirement
      ↓
Resource Requirement
      ↓
Resource Forecast

The same resource engine could support different domains while allowing different optimisation rules.

Resource planning may sit above scheduling

This distinction is useful.

Scheduling asks:

When should this known job occur?

Resource planning asks:

What capacity will we need to support the work we expect to emerge?

So:

Lifecycle / Production Forecast
        ↓
Resource Planning
        ↓
Work Generation
        ↓
Scheduling
        ↓
Execution

The scheduler remains important.

It operates within the capacity picture established upstream.

A working hypothesis

The proposition is not that work-order scheduling is unimportant.

It remains essential for short-term execution.

The question is whether resource planning should start much earlier.

A useful distinction may be:

Scheduling allocates resources to known work.

Resource planning anticipates the capacity that future asset behaviour is likely to require.

That difference becomes important when:

• production changes;
• major components approach intervention;
• shutdowns are planned;
• labour is scarce;
• parts have long lead times;
• contractor capacity is limited.

The earlier the organisation can see those constraints, the more options it retains.

Questions worth testing

How far ahead can maintenance-resource demand be predicted with useful accuracy?

Should lifecycle forecasts automatically create labour and parts forecasts?

How should qualifications, site access and competencies be represented in available capacity?

Should workshop bays, cranes, specialist tools and facilities be treated as capacity resources alongside labour?

How should probable and risk-based work be included without overstating demand?

Can production plans be tested against maintenance-resource capacity before approval?

When should a forecast resource requirement trigger procurement, recruitment or contractor engagement?

And perhaps the broader question:

If much of tomorrow’s maintenance demand is already visible in today’s asset condition, production plan and lifecycle forecast, why should resource planning wait for tomorrow’s work orders to be created?

The work order is where resource demand becomes executable.

It does not need to be where resource planning begins.