The 48-Hour Look-Ahead: Make Tomorrow’s Work Ready Before Tomorrow Arrives

At 06:55, eight people are standing at the gate. The programme says scaffold erection starts at 07:00. The supervisor is onsite. The truck has arrived.

Then the gaps appear. The latest drawing is awaiting approval. The load is missing components. Nobody has confirmed the access route. The required permit is still outstanding.

Three hours later, the supervisor is arranging alternative work.

This illustrative situation exposes a control failure: information about the blockers existed before the crew arrived. It never became a timely decision.

A 48-hour look-ahead moves that decision upstream. Instead of asking why the work could not start, it asks what could prevent the start, who will resolve it and when the evidence must be available.

What a 48-hour look-ahead actually does

A 48-hour look-ahead tests whether upcoming work can become a credible daily commitment.

Here, “48-hour” means the next two planned working days, reviewed at a fixed daily cutoff. Record the actual window start and end timestamps. Two working days can span more than 48 elapsed hours across weekends, holidays or shift patterns.

It fits into a simple planning hierarchy:

Long lead items still need earlier attention. The short review checks that earlier planning has produced the conditions the work needs.

Scheduled is not the same as ready

“Wednesday: erect 450 m² of scaffolding” contains assumptions about drawings, materials, skills, transport, access, predecessor work and permits.

Use a defined work package: a specific scope, location, planned start, committed finish and measurable output. Agree what counts as acceptable completion.

The eight readiness gates

Use the same gates in the review, register and daily plan. For each applicable check, record its acceptance evidence, verifier and required-by time. Explain any non-applicable gate.

1. Scope and information

Confirm the work boundary, quantities, current drawing or instruction, revision and unresolved technical questions. The evidence might be an approved drawing reference and a confirmed work-package instruction.

2. Predecessor work

Check that preceding activities are complete to the required acceptance standard, the area has been released and conflicting trades have cleared it. Record the handover or inspection that establishes this.

3. Materials

Verify the requirement, quantities, condition, shortages, pick and delivery arrangements. Distinguish ordered, received, picked, loaded and delivered: these are different states.

The test is whether this work package can receive its complete usable requirement at the necessary location and time.

4. Labour

Confirm named crew allocation, required competencies, supervision, absences, competing bookings and mobilisation time.

5. Plant, fleet and equipment

Allocate specific suitable assets. Check availability, relevant inspections, maintenance conflicts, tools and any lifting arrangements. Confirm timing with the responsible asset or fleet controller.

6. Access and logistics

Verify delivery windows, route restrictions, unloading arrangements, laydown space, workface access and conflicts with other activities.

7. Safety and permits

Confirm applicable work methods, required permits, competent people, inspections, PPE, isolations and critical controls through the responsible site process.

Identify which requirements can be verified in advance and which need a check under the conditions present at the start of work. Record start-time checks as explicit hold points, with an authorised verifier. A planning review cannot substitute for them.

8. Commercial and client release

Confirm the instruction to proceed, scope authority, applicable variation approvals and relevant contractual hold points. Use the approval route that actually governs the work.

Record who authorised the activity and the supporting reference.

Two decisions: commit the plan and authorise the start

At the planning cutoff, verify every prerequisite due before commitment. Checks that can only occur at the workface must already have a named verifier, an agreed verification time and a recorded hold point.

Immediately before execution, the authorised person verifies all applicable start conditions. Where a permit is required, work stays on hold until the permit is valid and its conditions are met.

Planning-ready does not mean authorised to start.

Do not turn this into “seven out of eight gates passed, therefore 87.5% ready”. One unresolved critical prerequisite can block the decision being made, regardless of how many other boxes are Green.

Use Green, Amber and Red consistently

GREEN — Planning-ready. Every prerequisite due by the commitment cutoff is verified. Any required start-time checks are explicitly arranged and recorded. The activity can enter the committed daily plan.

AMBER — Recovery required. A prerequisite remains open before its required-by time, with a named owner and a credible recovery path. Keep the activity in the look-ahead while the owner resolves it. It is not yet a firm daily commitment.

RED — Hold or resequence. A critical prerequisite is overdue, has failed, or has no credible route to resolution by its required-by time. Resolve, escalate or select ready alternative work.

At the workface, use a separate GO/HOLD decision. GO requires every applicable start condition to be satisfied. Green on yesterday's board cannot override today's HOLD.

Run the T-48 → T-24 → T-0 cycle

T-48: identify and assign

Two working days before execution, test the eight gates. Open a constraint for each unresolved prerequisite, with an owner, acceptance test and required-by timestamp.

Escalate immediately if the remaining lead time is insufficient.

T-24: verify and prepare the commitment

One working day before execution, replace “Is it sorted?” with “What evidence confirms it?”

Check drawing references, complete material picks, named assets, crew allocations and access confirmations. Verify permit prerequisites and arrangements for any required start-time checks.

At the agreed daily commitment cutoff, commit planning-ready packages. Retain the original plan and record subsequent changes. Prepare a ready fallback where an outstanding start hold point could prevent the primary activity.

T-0: authorise execution

Before work begins, verify the start conditions and changes since commitment. Record GO/HOLD, the responsible person and the time.

If conditions have changed, reassess. A prior plan commitment is never a reason to bypass a start requirement.

Work backwards from the start time

A Wednesday 07:00 start needs earlier operational deadlines. For example:

These times are illustrative. Calculate actual deadlines from loading, travel, approval, recovery and verification lead times.

Preserve the original required-by deadline, plus reasons and authorisation for revisions. Moving deadlines must not hide missed commitments.

Worked example: Wednesday scaffold erection

Assume an illustrative package at Site B: an eight-person crew, a Wednesday 07:00 start and 450 m² of defined scaffold scope, with an agreed finish time and acceptance criteria.

At Monday's review:

Overall status is Amber, assuming each open item has a credible recovery path before its deadline. If the drawing cannot arrive in time, the overall status becomes Red.

Create three separate constraints:

By Tuesday's commitment cutoff, accepted evidence closes all three. If the remaining gates also remain satisfied, mark the package planning-ready and commit it.

Where the site process requires permit issuance or checks at the start of the shift, retain those explicit start hold points. On Wednesday, record GO only after they are satisfied. Otherwise record HOLD and use the authorised alternative plan.

Understand the cost without inventing a saving

Consider a hypothetical combined attributable resource rate of R1,500 per hour for the people and assets actually immobilised by a readiness failure.

Three hours represents R4,500 of resource-cost exposure. If an identical event occurred twice weekly across 50 operating weeks, the scenario would total R450,000.

That is a scenario, not a measured result or an automatic cash saving. Separate additional expenditure, paid capacity that cannot be recovered and opportunity cost. Count each resource once. Account for useful alternative work and substantiate any additional recovery overtime or transport cost.

Measure the result against a baseline before claiming a financial benefit. The downtime-cost method explains how to separate resource exposure from additional expenditure.

Freeze requirements and record exceptions

Repeated changes after picking, loading and departure destabilise the plan.

Set a clear freeze cutoff for the relevant execution window. Keep a snapshot of the requirement IDs, quantities and destinations at that time.

Later changes can still be justified. Record the changed requirement, reason, requester, approver and operational effect. Distinguish changed frozen requirements from entirely new requirements added after the cutoff.

Emergency work, changed conditions, client instructions and incomplete forecasting need different responses.

Build a register that can carry the evidence

Start with a work-package register and a linked constraint list. One activity may have several blockers with different owners and deadlines; do not force them into one shared cell.

Use the register as the control record and link its event IDs to the daily site report. “Sorted” in a message cannot replace accepted closure evidence.

Measure five things consistently

Fix the reporting window and preserve the relevant baseline. For every percentage below, a zero denominator means N/A, not zero or 100%.

1. Ready Work %

Planning-ready packages at the review cutoff ÷ all packages in the defined look-ahead cohort at that cutoff × 100.

Record additions, removals and resequencing separately. Removing a blocked package must not silently improve the original cohort's score.

For a labour-weighted view, divide baseline planned labour-hours attached to ready packages by baseline planned labour-hours for the whole cohort. This shows how much planned labour is associated with ready work; it is not a probability of successful execution.

2. On-Time Constraint Closure %

Constraints due in the period with accepted evidence at or before their required-by time ÷ all constraints due in that period × 100.

Include constraints still open after their deadline. Use original deadlines for the baseline result, retain approved revisions separately and show older overdue constraints alongside the measure.

3. Committed Work Reliability %

Committed packages meeting frozen output and acceptance criteria by their committed finish ÷ all committed packages due to finish in the period × 100.

Include commitments subsequently held, cancelled or changed; record those reasons separately. Excluding them would rewrite the promise being measured.

A miss does not automatically prove a weak readiness gate. Investigate readiness, execution, scope change and changed conditions separately.

4. Late Change Rate

Unique frozen requirement IDs changed after cutoff and before the execution window closes ÷ requirement IDs present at freeze × 100.

Count each frozen requirement once even if it changes repeatedly. Report entirely new post-freeze requirements separately as a count. Mixing change events or new items into this percentage produces a different, potentially misleading measure.

5. Hours Lost to Readiness Failure

Record person-hours, plant-hours and vehicle-hours separately, with a documented readiness cause and start/end interval. Identify whether the resources performed useful alternative work.

Use consistent site delay codes to identify the recorded cause. Apply defined resource rates without double-counting.

Compare readiness with delivery carefully

Compare matched work packages and reconcile the findings with planned versus actual labour hours. Inspect the causes of misses and the decisions made during the look-ahead.

A shared trend does not prove causation. Work mix, capacity, weather, scope and site conditions can also change performance.

Keep the daily review focused

Aim for a ten-minute exception review with evidence prepared beforehand. Verification and problem-solving take additional time; larger workloads may need longer.

Use five steps:

1. Review Red packages and decide whether to resolve, escalate or resequence.
2. Review Amber constraints and the actions due today.
3. Check missed required-by times and owners without a credible recovery path.
4. Review changes since the freeze and their effects on resources.
5. Commit planning-ready work, identify start hold points and confirm ready alternatives.

End with decisions, owners and deadlines.

Implement it with the next ten activities

Define ten packages with output, timing and acceptance criteria. Apply the eight gates at the agreed cutoff.

Assign each constraint an owner, deadline and acceptance evidence. Preserve the committed plan. Verify start conditions and record the execution result.

After a week, inspect misses and refine the checks where failures occurred.

After 30 days, compare results with the baseline, allowing for changed workload and work mix. Check adoption, evidence quality and whether reviews changed decisions. This is a review point, not a promised improvement period.

The OperatorOS Principle

Do not discover tomorrow's problems tomorrow.

Bring the drawing problem into view before mobilisation. Confirm the material requirement before loading. Resolve access before unloading. Arrange permit prerequisites early and verify the required start conditions at the workface.

Unexpected events will still happen. The purpose is to reduce avoidable surprises from information the organisation already had.

Can you show why tomorrow's important work is ready for commitment, and what must still be verified before it starts?

Make that answer visible. Give every unresolved prerequisite an owner and a deadline. Let evidence determine the next decision.

Explore more practical operating systems in the OperatorOS Journal.