13The mission lifecycle

Delta V is not just a simulator; it's the whole life of a mission, from the first orbit sketch to a reviewed, flown, archived record. This chapter tours the desks and rooms that carry a mission through its life — every number below is computed by the same deterministic engine, so you can trust it and reproduce it.

Design it — the Flight Deck desks#

The Flight Deck rail groups its desks by the work you're doing:

• Design — Orbit, Maneuver, Target, Finite burn, Sequence, Vehicle, Requirements.
• Observe — Track a real satellite (SGP4/SDP4), Access, Coverage.
• Precision — Determine orbit, Conjunctions.
• Truth — Fidelity (how much to trust the number).

Two design desks are worth calling out:

Sequence chains coasts and burns into one executable mission and runs the whole thing through the engine in a single deterministic call. It seeds itself with the engine's own LEO→GEO transfer, so you can see a textbook Hohmann rediscovered, then edit it. Press ⌘Z to undo an edit.

Requirements (THE LEVELS) is the engineer's V&V table. Each requirement — "operate below 2,000 km", "keep eclipse duty under 40%" — carries a live verification link to an engine-computed value. Change the orbit and a requirement flips from ✓ VERIFIED to ✗ FAILED in place. Nothing is a stored checkbox; every verdict is re-derived from the current orbit.

Design the spacecraft — the Clean Room#

The Clean Room (/clean-room) is where you design the vehicle itself: a subsystem tree with live mass, power, link, and Δv budgets. The power budget uses the engine's real eclipse duty cycle (how much of each orbit is in Earth's shadow); the link budget solves Eb/N0 against range; the Δv budget uses the rocket equation. Edit any input and every budget re-closes — a 6U CubeSat design flips to "design closes ✓" the moment its solar array is big enough.

Weigh the options — Trade Studies#

A Trade Study (/glass-room/trade) sweeps one design variable — say altitude — and computes the consequences per candidate, then records the decision at the knee of the trade. It's the "why this orbit" table a reviewer expects, with every row an engine call.

Review it — the Glass Room & FLIGHT RULES#

Every mission walks a lifecycle: DRAFT → ON THE PAD → SIM → FRR → GO → FLOWN → SAFED. Entering SIM runs our own engine against the mission's claimed numbers before any human reviews it (auto-SIM) — the anti-cheat that only a deterministic engine makes possible. The Glass Room is the director's console: a review queue, a compare console for two forks, a budget desk (can this rocket fly this mission?), and Go/No-Go polls. Mission documents keep an append-only revision history (/m/[id]/rev/[n]).

Remember it — Mission Heritage#

The Mission Heritage room (/heritage) is the CANON plane: curated, cited, immutable-by-review records of missions that actually flew — Apollo 8 to Artemis I. Each record's key claim is verified live by the engine: Apollo 8's trans-lunar injection computes to 3,137 m/s against the published 3,140 — a 99.9% match. Missions beyond a two-body model (JWST's L2 halo, Artemis's distant retrograde orbit) are honestly gated with the reason.

The whole trip, once#

You can watch a complete profile end to end: open the Flight Deck and fly the Luna Sample Return (or share /deck?fly=luna-sr). It plays all ten phases — launch, TLI, lunar orbit insertion, powered descent, sampling, ascent, the first-robotic-docking rendezvous, trans-Earth injection, skip reentry — with real burn plumes, staging, a landing ring, and a docking target, and a reviewer panel at the side showing mission-elapsed time and the Δv ledger draining against its budget.

Ask about any of it — CAPCOM#

The AI copilot, CAPCOM, lives in a dock on every deck. It never estimates a number the engine can compute, and it cites only sources a human has reviewed. Pick one of seven console lenses (CAPCOM, FIDO, GUIDANCE, EECOM, INCO, FLIGHT, PAO) to reframe the same grounded answer — a flight dynamics officer and a public-affairs officer explain the same Δv differently, but from the same numbers. Every answer shows its confidence (high = engine-computed, grounded = corpus-cited) and the trace of tools it called.

See For engineers for the fidelity ladder and Fidelity & honesty for how we label what we know.