All systems operational UTC

The platform

Helion is a small set of tightly integrated services. We try to keep this page honest and current; if it disagrees with the live product, the product wins, and we update this page within a week.

01

Telemetry

One normalised state model across every bus. Sub-second stream-to-API. Replayable contacts, original raw frames retained.

02

Tasking

A single opportunity queue, ranked by mission value and fuel budget. Drag-and-drop re-tasking with diff and roll-back.

03

Anomaly

Watchdog rules run on every contact. Configurable response chains. Full incident timeline attached to a single URL.

04

Downlink

Ground-station orchestration across our network and yours. Handoff is automated; failover is automatic and visible.

05

Audit

Every command signed and replayable. Every operator action timestamped. Regulator questions answered in an afternoon.

06

API

GraphQL and gRPC. The same surface we use internally. No legacy XML, no "contact us for the enterprise version".

Supported buses

Full bus list →

6U CUBESAT

AAC Clyde

Sub-decimetre pointing, 1U payload bay, ideal for IoT and AIS.

Tier-112 in production

EO MICROSAT

York Space

Mid-resolution optical, dual-mode S/X-band, our most-deployed platform.

Tier-138 in production

HEAVY EO

Kongsberg

Sub-metre optical, full-motion steerable, four-mode SAR.

Tier-29 in production

SOVEREIGN BUS

Custom

Bring your own. We have shipped against 14 sovereign buses in the last 36 months.

Custom14 supported

How a pass actually goes

The view from the operator's chair, during a routine LEO contact.

It is 04:14 UTC. The on-call operator is asleep. The Helion anomaly engine has been watching the downlink since AOS, ninety seconds ago. Three watchdog rules fired: a slow drift in the X-axis gyro, a transient over-current on the propulsion bus, and a missed tasking deadline from a peer operator.

The engine resolves the first two by issuing a routine reset and a ground-station handoff. The third it escalates, opening a P1 ticket and waking a human with a single attached URL: the full incident timeline, with original telemetry, command log, and the AI-generated summary. The human is on the call twenty-three minutes later. The anomaly is already contained.

This is what we are building. Not a dashboard. A place where operators can stop running their own ground software.

What we will not build

Short list. More important than the long list.

  • No proprietary bus. We work with whoever you fly.
  • No "AI operator". The human is always in the loop. The engine escalates; it does not decide.
  • No data retention surprise. Customer telemetry is owned by the customer, full stop.
  • No per-seat pricing on the operator side. Per-satellite, per-contact, and per-task — that is it.
  • No vendor lock-in. We export the entire state model in a documented open format. Migration is a documented project, not a six-month reverse engineering.

What we will build next

The 2026 roadmap, in priority order. Quarters, not promises.

Q2 2026Anomaly-aware tasking (4.12 — live)
Q2 2026Sovereign cloud deployments (EU, GCC, ASEAN)
Q3 2026Federated tasking across operators (opt-in)
Q3 2026Direct X-band uplink integration
Q4 2026Multi-mission planning for SAR + EO fusion
Q1 2027Helion Edge: operator-side on-prem runtime