Case 3 Prototype First

Prototype First

Metro Real-Time Control Prototype

This is a technical feasibility checkpoint for critical infrastructure — not a consumer MVP sprint.

1. Problem

A technology company is developing a passenger access control system for a metro on a new domestic computing platform with a non-standard architecture and a custom real-time operating system (RTOS). No prior implementation of this platform exists in transport access control applications.

Three critical technical questions have no answers before the prototype:

  • Does the new processor deliver the deterministic response time required for metro turnstiles?
  • Is the custom RTOS compatible with hard real-time interrupt handling under peak load?
  • Is the computational capacity sufficient for peak-hour transaction throughput?

The project sponsor is considering the full project scope. The team asks to start. The question is: should the full engineering effort be committed before these questions are answered?

2. Choice

TA → PP Prototype only — initial delivery target is H1

The calculator uses the full lifecycle (Choice #1) for labor distribution, but the active delivery scope covers only TA and PP — the stages required to reach H1 (Prototype). Stages TP, WP, and IM are staffed at 1 FTE each — a deliberate signal that they are not included in the active delivery scope until the prototype checkpoint is reached.

3. Target Stage

Prototype Horizon H1 — checkpoint

After H1: if the prototype confirms feasibility — proceed to full cycle. If not — switch to proven hardware and standard stack. Decision based on fact, not faith.

4. Mapping Note

For this project, 5 functions were selected via the Function Mapping Procedure (FMP). Full function composition is available inside the calculator.

Technical Complexity Hard Real-Time Constraints
Architectural Complexity Real-Time Interactive Experience
Innovation Zero-to-One Breakthrough, New Computer Type, New Operating System
Standard Software Reuse ≤20% — Mostly custom (minimal reuse)
Why 5 functions produce 7 930 pd — the same function count as Case 1, but 2.2× more labor. Zero-to-One Breakthrough + New Computer Type + New Operating System act as a labor multiplier. Innovation is not free — and the calculator makes that effort explicit.

5. Report View

Engineering resource allocation: TA=5, PP=5, TP=1, WP=1, IM=1  |  Annual working time: 235 days/year per FTE

TP, WP, IM are held at 1 FTE — not included in the active delivery scope until the prototype checkpoint at H1.

Horizon Stage Product Stage Labor (pd) Team (FTE) Time from Start Status
H0 TA — Technical Assignment Requirements Baseline 923 5 0.79 yrs Active Scope
H1 PP — Preliminary Project Prototype ← checkpoint 755 5 1.43 yrs Active Scope
H2 TP — Technical Project MVP 923 1 5.36 yrs Frozen
H3 WP — Working Project Release Candidate 4 154 1 23.03 yrs Frozen
H4 IM — Implementation Production Release 1 175 1 28.03 yrs Frozen
Total (full lifecycle) 7 930 pd 28.03 yrs (frozen config)
Entry effort to checkpoint H1: 1 678 pd (TA 923 + PP 755) out of 7 930 pd total — 21% of the full calculated engineering effort. For 21% of the full calculated engineering effort, the project sponsor receives a binary answer: the new platform works for this application, or it does not.

6. Decision

Execute Phase 1 only: TA + PP — 1.43 years to prototype. The prototype is demonstrated on test equipment. After the demonstration, two scenarios:

  • Scenario A — prototype works: the new platform confirms feasibility. Full team is deployed on TP, WP, IM. The project continues on the new hardware with a proven technical foundation.
  • Scenario B — prototype fails: the hypothesis is disproven. The project switches to proven hardware and standard stack. Loss: 1 678 pd instead of 7 930 pd — 79% of the engineering effort is preserved.

The decision at H1 is based on a real artifact, not a promise. This is the calculated structure for a project with maximum technical uncertainty.

7. Engineering Feasibility Analysis

Prototype First

Committing full resources before the prototype means allocating full effort to an unverified hypothesis. The calculator separates the effort required to enter the unknown (21%) from the effort of building the full system (100%).

This is not caution — it is the calculated engineering route when the core technical assumption has no precedent. Phase 1 requires 1 678 pd. It returns a binary answer that is worth the full difference between 1 678 pd and 7 930 pd.

Calculated checkpoint structure: Phase 1 — execute H0+H1 (1.43 years, 1 678 pd). Checkpoint — prototype demonstration on test equipment. Phase 2 — expand to the full lifecycle only after a positive checkpoint result.

Delivery model: Full Turnkey