TraCSS Satellite Service: What It Does and What It Misses

In February 2026, NOAA opened the waitlist for TraCSS — the Traffic Coordination System for Space — a free government service that gives small satellite operators access to basic Conjunction Data Messages. For teams running CubeSats on shoestring budgets, it sounds like the answer to the monitoring problem. But a CDM is a data file, not a decision — and that gap is where most operators get stuck.

What Is TraCSS and Why Does It Matter?

For years, small satellite operators had two options for conjunction screening: pay a commercial provider like LeoLabs, or rely on the 18th Space Defense Squadron's publicly available screening services — a system designed for a different era of space traffic, primarily aimed at larger, better-resourced programs.

TraCSS changes that. NOAA's Space Traffic Management mandate — formalized through Space Policy Directive 3 — tasks the agency with developing a civil STM system to handle the growing commercial and civil space sector. TraCSS is the operational result: a government-run service that pushes Conjunction Data Messages directly to operators, at no cost.

This matters because the economics of CubeSat programs are brutal. A university team or early-stage operator often can't justify a $20,000/year commercial screening contract. TraCSS removes that barrier entirely. It's the government finally building infrastructure for the operators who need it most — the ones flying on 3U and 6U form factors with no dedicated flight dynamics team and a $0 operations budget.

The February 2026 waitlist opening is the public availability signal. This is real, it's coming, and every small operator should understand what they're getting.

What TraCSS Actually Gives You

A CDM (Conjunction Data Message) is a standardized data format — CCSDS 508.0-B-1 — that encodes the predicted close approach between two resident space objects. Here's a representative slice of what a CDM contains:

OBJECT                = OBJECT1
OBJECT_DESIGNATOR     = 25544
OBJECT_NAME           = ISS (ZARYA)
TCA                   = 2026-03-19T14:22:31.000
MISS_DISTANCE         = 412.0  [m]
RELATIVE_SPEED        = 14253.2  [m/s]
COLLISION_PROBABILITY = 1.47e-04
COLLISION_PROBABILITY_METHOD = FOSTER-1992
COVARIANCE_METHOD     = CALCULATED
...
CR_R, CT_R, CT_T, CN_R, CN_T, CN_N, CRDOT_R, ...

The key fields: object IDs (NORAD catalog numbers), TCA (Time of Closest Approach), miss distance in meters, a Pc (probability of collision) estimate, and a covariance matrix describing the uncertainty in each object's position and velocity.

That covariance matrix is actually the most valuable part of a CDM — it tells you how confident the tracking data is, and it feeds directly into Pc calculations. More on how those calculations work.

The format is machine-readable by design. It's dense. A human operator staring at a raw CDM text file is not going to make a fast, confident decision from it.

TraCSS Limitations You Should Know

This is not a criticism of TraCSS. It's designed as civil infrastructure — a data delivery layer — not an operator decision interface. Knowing its boundaries helps you use it correctly.

No visualization. You get CDM files. You don't get a plot of the conjunction geometry, a miss distance timeline, or a trajectory view. That's not TraCSS's job.

No trend tracking. A single CDM is a snapshot. Pc can evolve significantly over the days leading up to TCA as tracking data improves. TraCSS doesn't surface that trend — you'd need to manually compare successive CDMs across the screening cadence.

No operator-readable geometry breakdown. A CDM's covariance matrix encodes the radial, in-track, and cross-track uncertainty — but as raw numerical fields (CR_R, CT_T, CN_N...), not as an interpreted conjunction picture. Most operators need that geometry translated into something actionable: which axis dominates, what maneuver type would help, what the miss distance looks like over time. TraCSS delivers the data; interpreting it is on you.

Covariance quality varies. This is the honest reality of the current tracking catalog. For objects without recent high-quality observations, the covariance data in CDMs can be degraded or unrealistically tight. A low Pc estimate with bad covariance is not a clean all-clear.

Update frequency and latency. TraCSS targets screening at defined intervals, not real-time. There will be windows between updates where your knowledge of a developing conjunction is stale. For high-Pc events close to TCA, that matters.

Again — these aren't failures. They're the expected properties of an infrastructure service. The question is: what fills the gap?

The Decision Gap: Alert vs. Action

Getting a CDM is the easy part. The hard part is answering four questions:

1. Is this in the noise? Operators routinely receive CDMs for conjunctions where Pc is 1e-6 or lower. These are background events — real, but not operationally significant. Distinguishing signal from noise requires context, and raw CDMs don't provide it.

2. Has Pc been trending? A Pc that started at 1e-5 four days ago and is now 1e-3, with TCA in 36 hours, is a very different situation from a 1e-3 that just appeared with TCA in 10 days. Trend matters more than the current snapshot.

3. What does the geometry tell you about maneuver options? The radial component of a conjunction tells you whether a timing maneuver (in-track burn) is effective. The cross-track component tells you whether an out-of-plane burn is warranted. You can't get that from a Pc number alone. You need a full conjunction risk assessment.

4. What do you document? If you maneuver — or decide not to — that decision needs to be defensible for regulatory purposes. A raw CDM doesn't constitute compliance documentation.

This is the decision gap. It's not a technical problem. It's an operational one. And it's exactly where most small operators — the ones TraCSS is designed to help — have no tooling in place.

See your satellite's current conjunction picture → Run a free screening

How OrbVeil Complements TraCSS

OrbVeil isn't a replacement for TraCSS. It's an operator layer that sits on top of whatever screening data you have.

Here's what OrbVeil adds:

Independent daily screening. OrbVeil screens against 29,654 cataloged objects daily, completing a full propagation run in 9.6 seconds. This gives you a parallel screening track that doesn't depend on TraCSS update cadence. You get coverage even in gaps.

ML-based Pc classification. Raw Pc values are noisy. OrbVeil layers a binary alert classification model trained on historical conjunction data — AUC=0.924 — that separates actionable events from background noise. This is the signal-from-noise problem, addressed quantitatively.

Notably, the model's performance improves significantly when covariance data is available. We've measured a Spearman correlation of ρ=0.694 with covariance features versus ρ=0.469 without — a delta of 0.225. That makes TraCSS's covariance data more valuable, not less. When you feed TraCSS CDM covariance into OrbVeil's ML pipeline, you get better classification than either system provides alone. See our validation methodology.

RIC/RSW breakdown on every event. OrbVeil converts all conjunction geometry into radial, in-track, and cross-track components. This is what your team actually needs to evaluate maneuver options — not the raw covariance matrix.

PDF compliance export. Every screening event can be exported as a formatted PDF report suitable for FCC record-keeping. This matters for operators operating under the FCC 5-year deorbit rule — documented screening is part of demonstrating responsible operations.

Free for university teams. If you're a university CubeSat program, OrbVeil's full access tier is at no cost. Same data, same ML model, same reporting.

The Workflow: TraCSS + OrbVeil in Practice

This is a repeatable workflow any small operator can implement today:

Step 1 — Enroll in TraCSS. Sign up for the TraCSS waitlist at NOAA's Space Traffic Management portal. Free. No hardware required. You'll receive CDMs for your registered objects on each screening cycle.

Step 2 — Register your objects in OrbVeil. Submit your NORAD IDs (or your operator license identifiers) to OrbVeil. Daily screening begins immediately. You'll receive a daily report with all conjunction events above your configured threshold.

Step 3 — Cross-reference CDMs with OrbVeil reports. When TraCSS pushes a CDM for an event, check OrbVeil's report for the same conjunction. You now have two independent Pc estimates and OrbVeil's ML classification alongside it. Disagreement between the two is itself a signal worth investigating.

Step 4 — Use ML Pc and trend data for maneuver threshold. OrbVeil surfaces Pc trend over successive screening cycles. Use this — not just the snapshot Pc — to determine whether an event warrants maneuvering. Most operators set a threshold around 1e-4 for initiating maneuver evaluation, but trend data should influence that decision.

Step 5 — Export PDF for your compliance record. Any time you evaluate a conjunction — whether you maneuver or not — export the OrbVeil PDF report. File it. This is your documentation trail for FCC inquiries and demonstrates active monitoring under your license conditions. Full CubeSat collision avoidance monitoring workflow here.

Both services are free for university teams. There's no cost argument against running them together.

What to Expect as TraCSS Matures

TraCSS is an early-stage service. The February 2026 waitlist is an opening, not a fully operational deployment. Expect the screening cadence, coverage quality, and CDM delivery reliability to improve over the coming years as NOAA builds out the infrastructure.

That's not a reason to wait. Operators who build their monitoring workflows now — against real data, with real CDMs — will be well-positioned as the service matures. The operational habits matter as much as the tooling.

OrbVeil is designed to layer cleanly on top of any improvement in underlying data quality. Better covariance from TraCSS means better ML classification in OrbVeil. Faster CDM delivery means faster cross-referencing. The systems are complementary by design — not competitive.

Space traffic management is finally getting the civil infrastructure it needs. But infrastructure doesn't make decisions. Operators do.

TraCSS gives you the data. OrbVeil gives you the decision. Sign up free — universities get full access at no cost.

Apply for OrbVeil access →