Sequencing, sizing, and crew-assigning 100–500 tasks on a construction project — minimizing project duration under precedence and resource constraints (known in the literature as RCPSP).
In plain words
Sound familiar?
- The site supervisor calls every foreman at 7am to dispatch them — this round-up takes 30–60 minutes daily
- You run 3–6 projects in parallel and crew-sharing becomes a fight at month-end — 'which site am I on this week?' is a constant argument
- You can't pinpoint why a project slipped 2–4 weeks; every party blames a different cause
- Crane, concrete pump, or scaffolding sharing creates 'wait, let me finish' disputes 1–2 times a week
- When a supplier slips, you can't show numerically which downstream tasks shift — it's all gut feel
- Bad weather and 'how do I rebuild today's plan' eats half a day
- You told the client 'we'll deliver on date X' and you're slipping 2–3 months; penalty clauses or client churn follow
Why it matters
How it's solved
Technical depth
How it's solved
Technical depthOne-liner: Focus on the critical path — those are the tasks whose slippage drives the entire project late. Tasks off the critical path have slack — they can shift a day or two without hurting the finish date. Allocate the single crane or the single rebar crew to the critical-path tasks first.
What the software is really doing is this: the daily site-supervisor round-up that you draft on paper for weeks at a time, it lays out for 200–500 tasks in seconds and recomputes it every time something slips or changes. Three stages:
1. It describes the project. Each task — name, estimated duration, which tasks it must follow (precedence), which crews and machines it uses, which materials it needs. Crew inventory (how many formwork crews, rebar crews, concrete crews, where each is today), machine list (crane, pump, excavator), calendar (public holidays, weather forecast). The data flows from the bid/take-off system automatically, or is entered once into a clean table.
2. It computes the shortest feasible schedule. The software doesn’t try every possible sequence — for 200 tasks that’s mathematically impossible. Instead, it uses scheduling algorithms from operations research (a discipline that uses math and computing to solve business-decision problems) to take intelligent shortcuts: which task starts which day, with which crew and machine. A result comes back in minutes — all precedence constraints honored, no resource conflicts, minimum total duration. The critical path (the chain of tasks that drives the finish date) is shown explicitly; any slip on those tasks slips the project.
3. It reaches the field. The plan appears in the foreman’s mobile app as a daily task list: ‘07:30 5th floor formwork, 10:00 concrete pour on Block A, 14:00 rebar tie-down on Block B’. When a supplier slips or weather hits, the software recomputes only the affected section and a new schedule arrives in 15–30 seconds. With multiple projects, crew and machine assignments are optimized across projects dynamically.
It does not replace the supervisor’s judgement; think of it as a calculator that scales the 20-task plan they hold in their head to 500 tasks and computes the slip-propagation chain. The decision is still yours, but the numerical impact of every delay is visible.
Alternatives
Spreadsheet + the site supervisor's head
FreeFree
Who it fits: 1–2 projects, 30–50 tasks, small crew
- + Zero cost
- + Flexible — on-site adjustment is easy
- + No capex decision
- − Precedence and resource conflicts get missed at 100+ tasks
- − Critical path is invisible — you can't tell which slip moves the finish
- − Multi-project crew sharing can't be solved in someone's head
- − The promised delivery date has no mathematical basis
Local construction project management software
Enterprise$1,000–4,000 setup + $200–700/month (regional SMB pricing)
Who it fits: 3–10 projects, 100–300 tasks per project, stable crew structure
- + Local-language interface and support
- + Progress billing, take-off, and finance modules integrated
- + Site photo and daily-report workflow
- − Scheduling engine usually stops at Gantt drawing — no true RCPSP optimization
- − Critical path exists but resource-constrained automatic rescheduling is weak
- − Multi-project resource sharing is limited
International specialized project management software
Enterprise$30–100/seat/month subscription, or $25,000–200,000/year licence
Who it fits: Large contractor, highway/rail, 500+ tasks per complex project
- + Mature: RCPSP optimization, multi-project resource management, risk analysis fully supported
- + Critical path and Earned Value Management standard
- + Algorithms hardened over many years
- − High licence and consulting cost
- − Rollout takes 3–6 months
- − Local-language support limited; steep learning curve
Custom build on an open-source solver
Open SourceLicence free; 10–20 weeks of internal development, or $60,000–250,000 of consulting
Who it fits: Large contractor or a firm that builds the same project type repeatedly
- + No licence cost
- + Fully customizable to your work-item structure and standards
- + Runs in the cloud or on your own server
- − Requires technical capacity in-house
- − Ongoing maintenance is real work
- − RCPSP solver expertise is rare; building a team is expensive
Recommendation
Ask in the meeting
- Is the scheduling engine based on a true RCPSP algorithm, or just on Gantt drawing? How close to the optimum is it for 200 tasks and 10 resources?
- Is crew and machine sharing supported across multiple projects? How does the system decide which site a crew rotates to?
- Is the critical path computed under resource constraints (true resource-constrained critical path), or only classic PERT-CPM?
- When a task slips or weather hits, how fast does the plan recompute? How does it reach the field — app, SMS, phone?
- How does it integrate with progress billing, take-off, and finance systems? Do work items flow in automatically or are they entered manually?
- Is risk analysis (e.g. 'if supplier X slips, by how much does the project slip') done with Monte Carlo simulation or with simple assumptions?
- How do you structure the pilot — how many projects, how many weeks, what is the success bar?
- If we stop working with you, how do we get our task history, billing, and schedule data back? Is there a standard export format?
Technical details
Editor’s note
On site this problem is known as “the work schedule”, “the progress plan”, or “the site programme”. The academic name is the Resource-Constrained Project Scheduling Problem (RCPSP). It started with PERT/CPM in the 1960s and matured into RCPSP as crew and machine constraints were added — one of the longest-running fields in operations research. Without that vocabulary, in a software demo you cannot tell whether the “project management module” being pitched is a true resource-constrained scheduling engine or just a Gantt-drawing calendar.
The point most often overlooked in this segment: many products advertise “project tracking” but underneath they only do Gantt visualization plus manual drag-and-drop — meaning you do the planning and the software just renders it. A real RCPSP solver, when you tell it ’this task takes 3 days, there’s one formwork crew, the crane is free on that date,’ mathematically computes the shortest project duration. In any demo, ask the vendor to walk through a 30-task example with 4 crews and a crane constraint, and explain how the solver works and where the critical path lies.
A step-by-step path for an SMB
Stage 1 — Measure first, plan later. Over at least 2 projects, log four things:
- Estimated vs. actual duration per task line (what you said vs. what it took)
- Slippage cause (supplier, weather, crew shortage, planning error)
- Idle crew and machine hours
- Final project profit/loss vs. budget, in percentage
Without this baseline you can’t tell which software will deliver which result.
Stage 2 — Build your task-line and duration table. A typical residential block has 80–150 distinct work items (formwork m², rebar tons, concrete m³, MEP meters, finishing m²). Review each estimated duration and the variance you have seen. This is your knowledge capital — any serious vendor will ask for it first. Past data from 3–5 similar projects is gold.
Stage 3 — Pilot. Start with a single mid-sized project for its full duration (typically 6–12 months). Define the success criterion in writing, before the pilot: e.g. “delivery date variance under one week, idle crew hours down 25%.” If the bar is missed, the pilot ends — keep that exit right in the contract.
Stage 4 — Rollout. If the pilot lands, scale to all active projects over 3–6 months. Supervisor and foreman training runs 2–3 weeks; supplier-system integration adds 4–8 weeks if pursued.
Risks — what can go wrong
- Bad duration estimates. If your estimates are off, even the best schedule won’t hold in practice. Before the pilot, build an ’estimated-vs-actual’ table from past projects; a good RCPSP solver can also run with probabilistic durations.
- Site resistance. “Don’t let a computer tell me what to do” is a common reaction. In the pilot, walk through the results with the foreman; the software must transparently show which rule was applied and why a task is on the critical path.
- Supplier integration difficulty. When materials don’t arrive, the schedule is meaningless. The software must be able to pull supplier commitments in real time; if not, manual update load increases.
- Vendor lock-in. Software that stores progress billing, task history, and schedule data in a proprietary format makes migration hard. Put a clause in the contract: “We can export our data in standard open formats (CSV, XML, or similar) at any time, on request.”
Related cautionary lesson (will be linked once published): “A mid-sized residential contractor that dropped its project management software at month 11 — what they missed.”
A technical view of the solution method
This section holds what you’ll need when talking to a software team or a consultant. It is not what the supervisor sees on the daily screen — it is the engine behind the curtain.
The main approaches used for RCPSP:
| Approach | Typical scale | Solve time | Guarantees optimum? |
|---|---|---|---|
| CPM (Critical Path Method) | 100–1,000 tasks, no resource limits | Seconds | Yes (for the resource-unconstrained problem) |
| PERT | 100–500 tasks, uncertain durations | Seconds | Probabilistic expectation |
| MIP (mixed integer programming) | 30–100 tasks, resource-constrained | 1–30 minutes | Yes, given enough time |
| Branch-and-bound (specialized) | 50–200 tasks, resource-constrained | 1–15 minutes | Yes (mid-scale) |
| Metaheuristic (GA, SA, tabu) | 200–2,000 tasks | 30 seconds – 5 minutes | No (near-optimal) |
In practice: under 100 tasks, classic CPM may suffice; but with tight resource limits a metaheuristic gives a more realistic answer. For 200+ tasks, where MIP becomes intractable, a metaheuristic plus simulation is the standard combination.
Objective function choice changes the shape of the solution:
- Total project duration (makespan): “Finish early” — fits projects with contractual penalty clauses
- Resource utilization balance: “Keep my crews busy” — fits labor-cost-heavy contractors
- Cash flow (Earned Value optimization): “Owner payments on time” — fits cash-tight firms
- Late-delivery penalty + early-completion bonus: “Maximize net profit” — fits bonus-contract projects
Most real deployments use a weighted blend of all four.
Academic references
Listed in the sources block of this page’s frontmatter. RCPSP has been one of the most active operations research fields since the 1960s; current work focuses on multi-project scheduling, scheduling under uncertainty, and real-time recomputation. INFORMS Interfaces and the European Journal of Operational Research archive carry deployment case studies tied to real construction operations.
Sources
- Brucker, P., Drexl, A., Möhring, R., Neumann, K. and Pesch, E. (1999). Resource-constrained project scheduling: Notation, classification, models, and methods. European Journal of Operational Research, Vol. 112 — the canonical classification paper for the RCPSP field.
- Hartmann, S. and Briskorn, D. (2010). A survey of variants and extensions of the resource-constrained project scheduling problem. European Journal of Operational Research, Vol. 207 — the comprehensive modern review.
- Demeulemeester, E. L. and Herroelen, W. S. (2002). Project Scheduling: A Research Handbook. Kluwer Academic Publishers. The academic reference text.
- INFORMS Interfaces — case studies of operations research deployments in construction and project management. informs.org/Publications/Interfaces
Glossary
- RCPSP
- Sequencing hundreds of project tasks, sizing durations, and assigning crews under precedence and resource constraints.
- Critical Path
- The longest chain of dependent tasks running from project start to finish — the chain that drives the delivery date.
- MIP
- An optimization model where some decision variables are forced to be whole numbers (e.g. number of trucks, number of shifts).