Library / Projects / KEVOS® original book

Modular Projects by the Numbers

Standard modules, pilots, learning curves and short exposure windows: delivering one large project as many small repeatable ones

Plan and deliver a rollout as repeatable modules: outcome first, module boundaries, pilots, learning curves, exposure windows, cadence and an honest cost comparison.

10 chapters · About 105 minutes · Free to read · Intermediate project and manufacturing engineering practice

Start reading
KEVOS® LibraryPROJECTSModular Projectsby the NumbersKEVOS® ORIGINAL BOOK10 CHAPTERS

About this book

Many projects that look large are really the same job done many times: a dozen identical machines, forty substations, a hundred classrooms, twelve packing lines. Treated as one large bespoke project, each repetition is engineered afresh, installed slowly in place, and exposed for weeks to whatever goes wrong on a working site. Treated as a programme of standard modules, the design is done once, tested hard on a pilot, built faster each time by people who are learning, and installed in days. The modular route is not free: it costs more before the first unit pays back, and it fails if the module is chosen badly or the standard is allowed to drift.

This book works through that choice with numbers. It starts from the outcome the programme must deliver and works back to what should be standard, then covers interfaces and variants, iterating the plan with models and a pilot, proving the pilot with an acceptance test of the right length, and the learning curves that make repetition cheaper. It goes on to experience and team continuity, the risk of long exposure on site, rollout cadence and batch size, an honest comparison of the bespoke and modular options, and the change control that keeps a standard standard. Each chapter follows one running case, includes a plausible mistake and how it was found, and ends with an exercise and its worked answer.

It is for operations, engineering and project managers, automation and manufacturing engineers, and owners who plan multi-unit equipment programmes. It assumes spreadsheet arithmetic and logarithms at school level. The estimating and risk ideas it relies on are developed in Reference Class Forecasting by the Numbers, and this book links to it rather than repeating it.

The running case

Every chapter follows one fictional programme. This is an illustration. A snack-food manufacturer with three Australian sites packs product by hand at the end of twelve production lines: six at its largest site, four at the second and two at the third. Each line delivers 120 packs a minute into cases of 12, and two operators per shift hand-pack cases and stack pallets. The business plans to replace all twelve manual stations with automated end-of-line cells, each a case erector, case packer, checkweigher and robot palletiser. Each cell is expected to save three operator positions across two shifts, a net benefit of A$180,000 a year after its own running costs.

Two ways of delivering the programme are compared throughout. In the bespoke option, each cell is engineered for its own line and installed in place over about three weeks of disrupted production. In the modular option, one standard cell is designed once, proved in a pilot, built repeatedly in an integrator’s workshop, tested before it ships, and installed on its skids over a long weekend.

The business, its sites, its lines and every number in the case are invented. None describes KEVOS® work, a real business or a real project. All amounts are Australian dollars excluding GST.

How to use the book

Read Chapter 1 first, because the outcome it defines is the yardstick for every later decision. Chapters 2 to 4 cover design and proving, Chapters 5 and 6 learning and people, Chapters 7 and 8 the rollout, and Chapters 9 and 10 the comparison and the long-run discipline. Every calculation was checked by running it, and the KEVOS® tools linked in the chapters reproduce the learning-curve and cost arithmetic.

The book is one of a series of nine KEVOS® Library books on planning and delivering projects, written to be read together. Reference Class Forecasting by the Numbers supplies the outside view used here to judge each option’s risk, and Prototypes and Simulation by the Numbers goes further into models, prototypes and test sizes. First-of-a-Kind Technology by the Numbers covers the cost of being first, which a standard module avoids, and Project Teams and Contracts by the Numbers the contracts and teams that deliver a rollout. Front-End Planning by the Numbers, Stage Gates and Stopping Rules by the Numbers, Outcome-First Projects by the Numbers and Demand and Benefit Forecasts by the Numbers complete the series.

Source limits and professional responsibility

The ideas are established and the sources are listed with what each was used for. The case’s costs, rates and hours are teaching values, not market prices, and a real programme needs its own quotes, data and risk assessment. Packaging machinery and industrial robots are hazardous plant: their design, guarding, installation and commissioning are governed by work health and safety laws, the applicable codes of practice and standards, and the competent people responsible for them. This book teaches how to plan and compare delivery options. It is not a safety assessment, an engineering design or advice for any real installation.

Who this book is for

Operations, engineering and project managers, automation and manufacturing engineers, and owners of Australian businesses who plan multi-site or multi-unit equipment programmes and must choose between bespoke and standard designs.

What you will learn

  • Define a programme from its outcome, and decide which parts of it can become a repeatable module.
  • Use models, mock-ups and a pilot to find problems while they are cheap, and size an acceptance test run.
  • Fit, read and apply unit and cumulative-average learning curves, and estimate the cost of lost learning.
  • Quantify on-site exposure, set a rollout cadence, and compare bespoke and modular options on cost, time and risk.
  • Keep a standard design under change control so the benefits of repetition survive the rollout.

Inside this book

Chapters

  1. 01One project or twelve: starting from the outcomeDefine twelve packing cells by their outcome: 10 cases a minute, a pallet every 6 minutes, 36 positions and A$2.16m a year, then sketch bespoke against modular delivery.
  2. 02Drawing the module boundary: interfaces and variantsDecide what goes inside a standard packing cell, write a five-part interface specification, handle two case formats by settings, and measure 94% parts commonality.
  3. 03Iterating the plan: models, mock-ups and the cost of changeFind faults while they cost A$1,500, not A$31,000: an emulation model, a mock-up and a pilot cut the cost of 23 issues from A$713,000 to A$156,000, with one case-size miss.
  4. 04Proving the pilot: acceptance tests and run lengthPlan FAT, SAT and SIT for a packing cell, test 10 cases a minute and a pallet every 6 minutes, and size the run by the rule of three: 24 hours to show 1 stoppage in 8.
  5. 05Learning curves: unit and cumulative-average modelsCompare the unit and cumulative-average learning curves for twelve cells (13,191 h against 10,722 h), fit an 84.0% curve from six builds and forecast 870 h for cell 12.
  6. 06Experience, proven technology and keeping the teamProven machines versus first-of-a-kind integration, borrowed cost ratios of 1.046 and 1.496 at P80, choosing an integrator, and the 346 hours a crew change cost.
  7. 07Short exposure windows: installing in days, not weeksDisruption risk grows with days on site: 69.9% over 15 days against 21.3% over 3, A$228,000 against A$58,600 of site-related cost per cell, and a permit day nobody planned.
  8. 08Rollout rhythm: cadence, line of balance and batch sizeA cell every two weeks from a ten-week pipeline: five cells in progress, A$38,077 for each week the rollout slips, and why ordering all 11 cells early risked A$165,000 a change.
  9. 09Comparing bespoke and modular: cost, time and riskA$10.32m against A$8,446,131, break-even at the fourth cell, the last cell in week 92 or 68, borrowed cost ratios for the risk, and a comparison that left out A$520,000.
  10. 10Keeping the standard: change control and closing the programmeBaselines and change requests, A$15,120 per site deviation, a benefit review at 98.3% of A$2.16m, a cost ratio of 1.030, and a local change copied from cell 5 into cell 9.

Apply This in Practice

Use these KEVOS® tools to try the method from this reading with your own numbers.

References and source notes

The manuscript and worked examples are original KEVOS® writing. References inform particular concepts; their authors do not endorse this book. Web references were consulted on 28 October 2026.

  • Bent Flyvbjerg and Dan Gardner. How Big Things Get Done (2023). Supplied source, © Connaught Street Inc. and Bent Flyvbjerg, in copyright, read with its supplementary PDF. Used as a subject map for planning from the outcome backwards, iterative planning, experience, short delivery windows and modularity. No passages, chapter structure, stories or figures are reproduced.
  • Bent Flyvbjerg, Harvard Business Review 99(6), 58–63. Make megaprojects more modular (2021). Background on repeatable units, learning and speed as the argument for modular delivery.
  • Atif Ansar and Bent Flyvbjerg, Oxford Review of Economic Policy 38(2). How to solve big problems: bespoke versus platform strategies (2022). Background on the contrast between bespoke, one-off delivery and platform strategies built from standard parts, used to frame Chapters 1 and 9.
  • T. P. Wright, Journal of the Aeronautical Sciences 3(4), 122–128. Factors affecting the cost of airplanes (1936). Origin of the learning curve. Wright's model is stated for the cumulative average cost per unit; Chapter 5 sets it beside the later unit-cost form and shows how far the two differ.
  • Carliss Y. Baldwin and Kim B. Clark, MIT Press. Design Rules, Volume 1: The Power of Modularity (2000). Background on modules, visible design rules and hidden module internals, used for the interface discussion in Chapter 2.
  • Donald G. Reinertsen, Celeritas Publishing. The Principles of Product Development Flow (2009). Background on cost of delay and batch size, used in Chapters 8 and 9.
  • James A. Hanley and Abby Lippman-Hand, Journal of the American Medical Association 249(13), 1743–1745. If nothing goes wrong, is everything all right? Interpreting zero numerators (1983). Source of the rule of three: with no events in n trials, the 95 per cent upper bound on the event rate is about 3/n. Used to size the pilot acceptance run in Chapter 4.
  • International Electrotechnical Commission. IEC 62381 Automation systems in the process industry: factory acceptance test (FAT), site acceptance test (SAT) and site integration test (SIT) (2024 (edition 3.0)). Named by number and scope as the international standard for planning acceptance tests of automation systems. Its checklists are not reproduced; the test plan in Chapter 4 is the book's own.
  • Safe Work Australia. Model Code of Practice: Managing the risks of plant in the workplace (current edition). Named as the national model guidance for people who design, install, commission and use plant such as packaging machinery and industrial robots. Check with the local regulator whether it has legal effect in a given state or territory.
  • Standards Australia and Standards New Zealand. AS/NZS 4024 Safety of machinery (series) (parts and editions as called up). Named by number and scope as the series covering machinery safeguarding and risk assessment. No values or clauses are quoted.

Continue exploring

← All Library books