F-35 JPO Digital Engineering
Unclassified · Draft, 2026-08-14
45 minutes · the room

Run of show

Beat by beat: what runs live, and the single point each beat exists to make. Read Positioning and the Demonstration outline first — this page is the script for the room, not the argument.

1The 45 minutes

0–3
Their own findings slide, read back to them. Stove-piped development, insufficient verification, limited visibility.
We are answering the problem they stated, not the one we sell to.
3–6
The MSP-2 article. DMSMS-driven, their words, their project shape.
This is your problem, scoped to one card.
6–10
Run the baseline. It abstains. 15.2% coverage, no readiness figure reported.
A readiness number you cannot audit is worse than none — this is the tone for everything after it.
10–16
Stage 1. Requirements with verification methods assigned; the plan generated, not authored.
Verification plan automation, from their own REMEDE slide.
16–22
Stage 2. Run check_criteria.py, showing the freeze timestamp, the authority, and basis: drafted, not elicited. Then run the trade.
The part every vendor skips. Showing the freeze and its caveat is stronger than implying the weights are the customer's.
22–30
Stage 2b. Flotherm finds Candidate A at 110.12 °C against a 105 °C limit, measured. Say plainly that HyperLynx and Simcenter 3D were not run.
The multi-mode column. Nobody else in this competition can fill it.
30–36
Stage 3. Questa coverage and formal, then the OFP model in the loop via SystemC/TLM co-simulation. Name Innexis VSI as the product that scales this to system-of-systems, and say plainly it is not what just ran.
Parity on the incumbent's ground, deliberately underplayed. The reframe from "whose emulator is faster" to "who integrates before the hardware exists" is the point.
36–40
Re-run the readiness record live. The gate moves to NOT READY on the thermal failure. Show the discarded record and the unverified list.
Computed from evidence, cites objects that re-fetch, abstains where it must.
40–43
The four gaps, named. Stage 5 and 6 are where their money is and where our map is emptiest.
Credibility. This audience has commissioned reviews that found people overstating maturity.
43–45
Their closed-loop slide, with PLM at the top of it.
We sell that box. The vendor whose diagram it is does not.

2Talk track, the three moments that carry the room

Opening the baseline (min 6–10)

"Before we show you anything that passes, we're going to show you something that refuses to answer. This is the incumbent card, run through the same engine, on five simulation-only records. It comes back abstained — 15.2% coverage, no readiness number. That refusal is the point: a number you can't audit is worse than no number, and everything you see for the next half hour is held to that same rule."

The freeze, before the trade (min 16–22)

"These eight criteria were frozen at 00:45 UTC yesterday, before a single piece of Stage 2b or Stage 3 evidence existed — the tool prints the timestamp on every run and refuses evidence dated earlier. And we'll say the honest half out loud: these weights were drafted by our account team, not elicited from your stakeholders. What's proven is that the criteria came before the evidence. What's not proven is that they're your priorities yet — that's next quarter's conversation."

The gate moving live (min 36–40)

"Watch the gate. Before Flotherm ran, this candidate showed CONDITIONAL. We just fed it a real thermal result — 110 degrees against a 105 limit — and the record recomputes in front of you: NOT READY. Nobody edited a slide. The engine read a solver log and changed its own answer."

3What we will not claim in the room

Written down so nobody improvises past it under questioning.

  • Twenty-two of the thirty process skills behind this workspace have never been run against live data. They are specifications, said as specifications.
  • The Stage 2 criteria weights are ours, drafted for review, not elicited from JPO stakeholders.
  • Innexis Virtual System Interconnect (the standalone, SoC-scale product) has not produced any number shown today — Architect Explorer and Questa SystemC/TLM produced what's shown, and both are named as what they are, not as VSI.
  • 6.4.10–6.4.13 (HITL through sustainment) have no confirmed Teamcenter module. This is a roadmap conversation.
  • HyperLynx and the Simcenter 3D vibration case are proposed, not run, against this article.
  • No F-35 program data of any kind appears anywhere in this demonstration. MSP-2 is fictional.

4If time is short

Cut Stage 3, not Stage 2b. Cutting the emulation/co-analysis demo costs a beat on ground where we do not win outright anyway. Cutting multi-mode costs the argument — it is the one column a pure-EDA competitor cannot fill at all.

The readiness engine and its data are the floor: nothing below Stage 1 (requirements) and the frozen criteria can fail on the day, because both are already built, tested, and independent of the demonstration tier.