Production scale-up ยท Evidence guide
How do you scale drone production without hiding the constraint?
Define the target as accepted units of one controlled configuration in one stated period. Then connect started-unit demand, material releases, constrained resources, trained work, integration, test, acceptance evidence and cash timing. Multiplying a pilot-lot quantity does not prove that any of those systems can support the next rate.
Sources reviewed 2026-09-17. This guide translates public NASA implementation, integration, verification, transition and configuration-management guidance into a UAV ramp evidence screen. Apply the controlling customer, regulatory and organizational requirements.
| Gate | Question | Evidence to bind | Source basis |
|---|---|---|---|
| Configuration and demand basis | Which released configuration, accepted-unit quantity, mix and delivery period are being scaled? | Controlled product baseline, effectivity, demand ranges and change assumptions. | NASA-CM, NASA-IMPLEMENT |
| Started-unit and yield model | How many starts and rework cycles are needed to deliver the accepted quantity? | Measured yield, scrap, rework and acceptance definitions tied to the intended process. | NASA-IMPLEMENT, NASA-VERIFY |
| Material and supplier release | Which long-lead parts, raw materials and outside processes must be released, and when? | Dated availability, approved-source status, lead times, allocation and change control. | NASA-IMPLEMENT, NASA-CM |
| Flow and enabling products | Which station, fixture, machine, software or facility constrains the period? | Work content, sequence, queues and readiness for production and integration enabling products. | NASA-IMPLEMENT, NASA-INTEGRATE |
| People and standardized work | Which trained roles and released instructions are required at each rate step? | Staffing by skill and shift, training evidence, procedures and decision authority. | NASA-IMPLEMENT |
| Integration and test capacity | Can assembly, software load, calibration, test, retest and review support the same accepted rate? | Integration sequence, equipment and test-cycle capacity with interface and environment coverage. | NASA-INTEGRATE, NASA-VERIFY |
| Acceptance and transition | What evidence authorizes each unit or lot to leave production? | Verification results, discrepancies, approvals, as-built records and transition responsibilities. | NASA-VERIFY, NASA-TRANSITION |
| Cost, cash and phase gate | What is committed before the next rate step, and what observation stops the ramp? | Period-matched material, labor, tooling, equipment, inventory and cash timing assumptions plus an owner and decision date. | NASA-IMPLEMENT, NASA-TRANSITION |
Start with accepted output, not order quantity
Translate the accepted-unit target into starts using measured yield and rework assumptions. Compare that demand with each period constraint, including integration and test. If a single station can process fewer conforming units than the target requires, the plan has identified a gap, not a committed rate.
Advance through observed rate gates
- Freeze the configuration, period, mix and acceptance definition for the decision.
- Run the intended flow at a bounded rate and record starts, yield, rework, queue time, test time and accepted output.
- Compare the observation with the model and name the current constraint.
- Fund and verify only the next required change, preserving configuration and evidence lineage.
- Increase the rate only when the prior gate's acceptance and recovery conditions are met.
What this screen cannot establish
It does not establish a supplier's current capacity, a universal pilot-to-production threshold, a delivery schedule, market price, learning curve, financing need or flight approval. Those require dated evidence for the intended facility, product, quantity and period.
Download the scale-up gates JSON. Use the existing local ramp model.