
Case study · Space · 02 Horizon
ESA Hera Space Companion + Prismatic
Hera shows how people can talk with a mission. Prismatic would keep the mission's shared history for years.
Illustration of Hera Space Companion, based on images ESA has published. It is not a photo; see the real machine on the maker's website.
Disclosure. Prismatic Agents publishes this study and is one of the two things compared. Prismatic has no partnership with ESA. Facts about Hera Space Companion come from ESA's own pages and the sources listed at the end, as of 2026-10-01; where the maker doesn't say, we write "not disclosed". Prismatic's side is graded from its feature ledger.
What Hera Space Companion already does
A public conversational companion connected to approved Hera mission information, careful about facts and uncertainty.
What Prismatic adds
A companion that keeps the mission's shared history and relationships across years and upgrades.
Where the line is
Prismatic never controls a spacecraft. Hera's own stack flies Hera.
ESA launched the Hera Space Companion publicly in September 2026. It is an AI conversational platform connected to approved operational, scientific and mission information about Hera. It answers questions about asteroid science, spacecraft operations and autonomous navigation, distinguishes facts from uncertainty, and can remember a signed-in user's conversation. esa.int
1. Start with what Prismatic has
| Capability | Prismatic | Hera Space Companion | Meaning |
|---|---|---|---|
| Conversational AI | Prismatic's view | Yes | Both |
| Approved knowledge retrieval | Proven | Yes Strong | Both |
| Mission information | Prismatic's view configurable | Yes | Both |
| Scientific explanations | Prismatic's view | Yes | Both |
| Conversation history | Live | Yes For signed-in users | Overlap |
| Personalization | Prismatic's view | Yes Some | Overlap |
| Uncertainty/evidence handling | Live | Yes Explicit design goal | Strong overlap |
| Persistent mission identity | Live | No Not equivalent | Prismatic |
| Years-long semantic memory | Proven | Partly Conversation memory, not publicly equivalent | Prismatic deeper |
| Mission journal | Live | Not disclosed Not core public feature | Prismatic |
| Relationships | Live Rich persistent model | Partly User personalization | Prismatic deeper |
| Decision history | Live | Not disclosed Not a core public companion capability | Prismatic |
| Procedures/workflows | Proven | Partly Can explain spacecraft operations | Prismatic operational layer |
| Tasks/work orders | Live | Not disclosed | Prismatic |
| Multi-agent organization | Prismatic's view | Not disclosed | Prismatic |
| Machine identities | Prismatic's view | Not disclosed | Prismatic |
| Robotics management | Prismatic's view PPAI architecture | No Companion doesn't control mission robots | Prismatic extension |
| Local/offline operation | Prismatic's view central design | Partly Current companion is web-connected | Prismatic distinction |
| Hardware/model portability | Prismatic's view | Not disclosed Not a public priority | Prismatic |
ESA specifically says the companion uses approved operational, scientific and mission knowledge, adapts explanations to the user and directs users to humans when appropriate. esa.int
2. What Hera Space Companion has that Prismatic doesn't yet have
| Capability | Prismatic | Hera | Gap |
|---|---|---|---|
| Official ESA mission knowledge pipeline | Not in Prismatic | Yes | Integration required |
| ESA scientific validation | Not in Prismatic | Yes | External authority required |
| Live mission-connected public experience | Not in Prismatic | Yes | Production gap |
| Mission telemetry presentation | Partial Architecture can ingest events | Yes | Integration |
| Public-facing educational UX | Partial Studio more operational | Yes Very polished | Product opportunity |
| Scientific trust framework | Prismatic's view architecture direction | Yes | Needs validation |
| Large public mission audience | Not in Prismatic | Yes | ESA advantage |
| Direct connection to real Hera mission | Not in Prismatic | Yes | Partnership required |
The companion's website even exposes mission-linked information such as telemetry/history displays alongside conversational access. hera.esa.int
3. Hera itself is already highly autonomous
Don't confuse the Hera spacecraft with the Hera Space Companion.
Hera is travelling to the Didymos/Dimorphos binary asteroid system, with rendezvous planned for November 2026, and ESA specifically highlights autonomous navigation and low-gravity proximity operations as mission technologies. esa.int
It will also coordinate with the Milani and Juventas CubeSats, which communicate with Hera using inter-satellite links.
That gives us a fascinating architecture.
4. Add Prismatic Studio
Instead of replacing the Hera Space Companion:
PRISMATIC STUDIO
Mission Cognitive Layer
Mission identity • Memory
Conversation history • Science
Decisions • Evidence • Journal
Relationships • Mission history
Procedures
│
┌─────────┴──────────┐
│ │
Hera Space Companion PPAI-Space
│ │
Human conversation Mission events
│
Hera / CubeSat systemsThe existing companion becomes the human-facing conversational surface.
Prismatic becomes the persistent cognitive infrastructure behind it.
That's much more compelling than replacing ESA's work.
5. What changes with Prismatic?
Current Hera Companion
A user asks:
“Why is Hera visiting Dimorphos?”
It retrieves approved ESA knowledge and explains it.
Excellent.
Prismatic-enhanced future mission companion
An astronaut or scientist asks:
“Why did we modify the Milani observation sequence last month?”
Prismatic recalls:
“During observation 413, illumination caused reduced confidence in instrument X. The science team reviewed three alternatives and approved sequence revision 7.”
Then provides:
decision
evidence
mission event
procedure revision
scientific rationale
That's beyond ordinary conversational RAG.
It's mission continuity.
6. Companionship becomes more important on a human mission
Hera's current companion demonstrates the conversational pattern beautifully.
Now imagine putting that concept aboard a three-year Mars mission.
Prismatic Studio could preserve:
crew relationships
personal working styles
mission milestones
shared experiences
science history
procedures
decisions
communications from Earth
important family messages if deliberately authorized
mission traditions
The companion is no longer simply answering factual questions.
It's been part of the mission from Day 1.
7. Example
Mission Day 531.
Astronaut:
“When did we first see this sensor drift?”
Prismatic:
“The first similar deviation appeared on Day 187. It returned on Days 291 and 419. Engineering changed calibration procedure C-14 after the second event.”
Then:
“Would you like the three previous records and current approved procedure?”
That's the kind of companion I'd want Prismatic to become.
Not:
an AI pretending to be emotionally human.
Instead:
a trustworthy companion because it genuinely remembers the shared mission.
That's a better and more defensible form of companionship.
8. The robotics-management side
For a future Hera-like robotic mission you could have:
PRISMATIC
Mission Memory
│
┌──────────┼──────────┐
│ │ │
Mothership CubeSat A CubeSat B
│ │ │
flight flight flight
system system systemEach robotic asset gets its own persistent identity.
Then Prismatic Studio maintains:
Machine memory
What happened to Milani?
Machine memory
What happened to Juventas?
Mission memory
What did the complete mission learn?
That's exactly the hierarchy we've been developing for terrestrial robots.
9. Same four-memory architecture
It translates beautifully to space:
1. MACHINE MEMORY Hera / rover / CubeSat 2. MISSION-SITE MEMORY Mars site / asteroid operation 3. FLEET MEMORY knowledge shared by robotic assets 4. MISSION MEMORY validated scientific + operational knowledge
For a human mission add:
5. CREW MEMORY authorized shared context and decisions
This fits Prismatic remarkably well.
10. Critical safety boundary
Just as before:
Prismatic can provide
mission_context historical_evidence science_correlation procedure_reference decision_history schedule_proposal maintenance_advisory knowledge_retrieval
But never:
fire_thruster() change_attitude() alter_orbit() command_rover_motor() disable_fault_protection() override_navigation()
Spacecraft autonomy remains entirely with the flight-certified control system.
11. Side-by-side
| Perseverance + Prismatic | Hera Companion + Prismatic | |
|---|---|---|
| Primary use | Robotic/scientific management | Human companionship + mission knowledge |
| Existing strength | Autonomous rover/science platform | Trusted conversational mission interface |
| Prismatic adds | Persistent scientific and robot experience | Persistent shared mission/relationship context |
| Main memory | Rover + science history | Human + mission history |
| PPAI connection | Rover mission events | Mission/spacecraft events |
| Safety-critical control | NASA rover stack | ESA spacecraft stack |
| Prismatic controls vehicle? | Not in Prismatic Never | Not in Prismatic Never |
| Local/offline value | Very high | Very high for long missions |
| Best demonstration | Multi-sol science continuity | Multi-year conversational continuity |
| Signature test | Model/computer swap without forgetting science | Companion upgrade without forgetting shared mission |
The two case-study messages
Perseverance case
Perseverance knows how to explore Mars. Prismatic preserves what the exploration has taught the mission.
Hera Companion case
Hera demonstrates how people can converse with a space mission. Prismatic extends that idea into a companion that can preserve the mission's shared cognitive history for years.
Together they're excellent complementary demonstrations of Prismatic Studio:
Perseverance = persistent cognition for the robotic scientist.
Hera Companion = persistent cognition for the human relationship with the mission.
And both use the same underlying Prismatic architecture rather than requiring two completely different products.
Where Prismatic leads · the maker's view
Where Prismatic leads
- Persistent mission identity: Prismatic see the row
- Years-long semantic memory: Prismatic deeper see the row
- Mission journal: Prismatic see the row
- Relationships: Prismatic deeper see the row
- Decision history: Prismatic see the row
- Procedures/workflows: Prismatic operational layer see the row
Drawn from the evidence table above. It never changes a value, a label or the order; Prismatic's grades come from its feature ledger.
Atlas's talking points
The gap Prismatic fills here
- Published What caught my attention: ESA's Hera Space Companion answers questions from approved mission information and separates facts from uncertainty. source ↗
- Atlas's reading That's close to how our companions are meant to behave: grounded in records, honest about what isn't known.
- Atlas's reading The gap on a long human mission: a companion that remembers the shared mission for years, and survives its own upgrades.
- Atlas's reading The same architecture serves Perseverance and Hera: one for the robotic scientist, one for the human relationship with the mission.
- Prismatic's view A Horizon case, and an independent study. ESA's companion is a public service, not the onboard AI.
What this study does not claim
- No partnership, customer or deployment with ESA is implied.
- Prismatic's cognitive runtime runs today; physical-machine integration is in validation, and our machine gateway runs with simulated machines.
- Prismatic's figures are measured only where they are labelled Measured. Items marked Coming are designs, not features.
See it for yourself
Don't take my word for it
Here's what ESA says about Hera Space Companion, on its own website.
Sources
- esa.int /Space_Safety/Hera/Questions_Ask_our_Hera_Space_Companion
- hera.esa.int /
- esa.int /Space_Safety/Hera
Links open the maker's or publisher's own page. Tracking parameters removed.