It begins
Registered gto.space and opened the notebook for Garage to Orbit: a public attempt to understand what it really takes for an individual to build and fly a small satellite to orbit by rideshare.
Build log
The record starts before the hardware.
The useful story is the sequence of choices, constraints, tests, costs, and dead ends that make an orbital attempt credible.
Phase 1 architecture locked + Research Library
Completed bottoms-up mission definition for a Starship-deployed independent free-flyer (~50–150 kg wet mass, microsat / ESPA-class). Hosted-on-Starship remains an evaluated alternative. Baseline: 2× Starlink Mini Laser optical terminals for high-rate data (public claims up to ~25 Gbps; real programs on 60–500 kg buses including Muon Space Halo and Starcloud) + dual high-definition Earth-viewing cameras (1080p streaming baseline at 8–15 Mbps, grounded in flown missions such as NEMO-HD and Sen 4K on 16U). Orbit locked at 485 km altitude, 53° inclination for laser geometry and operations. Propulsion for collision avoidance and 5-year deorbit (no smallsat exemptions under FCC rules and NASA ODMSP/Conjunction Assessment Handbook). Full source library with master index and 100% traceability for every claim (NASA SmallSat SOA 2026, Starlink shell data, regulatory documents, vendor precedents). Decision 0004 and three single-source assumptions files captured internally. Public research page now surfaces the methodology and high-level synthesis; detailed working bundle remains in controlled internal access.
First-pass power and thermal
A first-pass power and thermal study is in for the locked 485 km / 53° orbit. Cameras-only closes. That means the power budget works if we fly cameras and skip the lasers for now. Dual-laser operation wants a deployable 250–350 W class array, not a small body-mounted bus — a folding solar wing in that class. Mini Laser size, weight, and power remain vendor-gated; nothing was invented to fill that gap. The compliance map is next. This is still an attempt, not a launch announcement.
First-pass compliance map
A first-pass regulatory map is in. If the payload images Earth's surface, NOAA contact is the first gate. The FCC licenses radio frequencies, not optical wavelengths. Five-year post-mission disposal was already required from Phase 1; 485 km is not a free ride. The printed smallsat track is not a schedule. It is a paperwork path, not a calendar. This is a high-level order of operations, not a legal memo and not a claim that the mission is licensed. Still an attempt, not a launch announcement.
First-pass bus interface
After separation we own the bus — power, thermal, tracking, and the five-year disposal showing. The live rideshare book is the August 2026 Payload User's Guide. The 50–150 kg class fits ESPA. ESPA is a common rideshare adapter ring; this mass class is built to fit it. Dual-laser operation wants a configured deployable wing; public ESPA-class products such as MMA rHAWK and FlexArray sit in that class, and Redwire ELSA's published floor is short of that row. Hosted Starship watts remain unpublished. This is still an attempt, not a launch announcement.
First-pass ADCS and propulsion
A first-pass attitude and propulsion study is in. HD Earth video is a tenth-of-a-degree body-pointing problem; public optical terminals do the fine pointing inside the head. Mini Laser pointing remains vendor-gated. A reaction-wheel suite in the XACT class is the typical close at 50–150 kg; that is a small pointing kit in that class, not a purchase. Control-moment gyros are not a product row at this mass. 485 km still sits in NASA's 400-versus-500 km decay gap. First-pass close is a ~50 m/s class 1-N pack — hydrazine or green monopropellant — not cold gas alone and not a drag sail as a collision-avoidance actuator. That pack is a small one-newton thruster with about 50 m/s of push, mostly to meet the five-year deorbit rule. No yearly collision-avoidance budget until someone runs conjunctions at 485 km / 53°. This is still an attempt, not a launch announcement.
First-pass TT&C and ground
A first-pass command and ground study is in. Optical terminals carry the data; they are not the command path. A dead laser still needs radio for collision avoidance and the five-year disposal burn. Typical close is S-band on the 2025–2110 MHz uplink / 2200–2290 MHz downlink pair, with a commercial ground-station network — KSAT Lite or Leaf — rather than a backyard dish first. Amateur UHF is not the primary path. In short, the lasers move the video; a commercial S-band radio still has to fly the satellite. A radio package still has to name transmit and receive frequencies, emission designators, EIRP, earth-station coordinates, a stop-buzzer, ITU, and the debris showing — optical stays in the narrative. This is still an attempt, not a launch announcement.
First-pass cost floor
A first-pass cost floor is in. The locked bus is mostly quote-only: the array, attitude suite, propulsion pack, and 2026 ground-network minutes do not print a unit price. The public Falcon SSO catalog is an SSO anchor only; 53° / Bandwagon has no printed dollar; the Starship smallsat ticket is missing. Those printed prices are for sun-synchronous orbits, not this 53° path, so they are not our launch ticket. NOAA remote-sensing contact is free. A Part 5 experimental or STA filing is $140, not the older $90. Next is RFQs and first hardware, not more architecture. This is still an attempt, not a launch announcement.
Executable baseline from Q1–Q6
The six first-pass maps are now one executable picture: what is locked, what closes, what stays gated, and what comes next. Next is RFQs and first hardware — not more architecture. Radio and remote-sensing paperwork wait for an explicit go. Conjunctions and first-hardware research can run in flight. This is still an attempt, not a launch announcement, and not a claim that the mission is licensed.
Conjunctions first-pass
A first-pass conjunctions check is in for the locked 485 km / 53° orbit. 485 km is a congested band. We do not have a cited collision-avoidance rate for this shell, and we did not invent one. A published 1–3 avoidance-maneuvers-per-year number (Funke) is for a different, higher polar orbit (~700–800 km sun-synchronous), not 485 km / 53°. The ~50 m/s class 1-N pack is still the first-pass close — post-mission disposal dominates the tank. Next is a real conjunction run, not another paper. This is still an attempt, not a launch announcement.
First-hardware first-pass
A first-pass look at what can actually be bought this week is in. Nothing that closes a locked flight need is add-to-cart. A cheap garage surrogate bench exists for cameras, encode, and a few flatsat coupons; it does not close S-band, the attitude suite, the 1-N pack, the deployable array, or a space camera. Shop pages that print euros are still not a cart. Class-close stays RFQ. The real flight parts still need a quote. A web shop cart does not close them. This is still an attempt, not a launch announcement.