Show the test.
Label the claim.

This page exists to be checked. Every activity has a status, every metric has a target or a measured result, and nothing unfinished is presented as finished.

Completed items marked green Test targets marked grey Failures documented, not hidden

Safety by design

Every risk has a control. Every control has a fail-safe state.

Safety is treated as a property of the architecture, not a feature. For each identified risk there is a primary control and a defined state the system falls back to if something fails.

RiskSystem responseFail-safe state
Obstacle suddenly enters the pathGraded deceleration; emergency stop if the safe gap is breachedFull stop before contact, brake held
Tracking target lostShort re-acquisition window; hold position if not recoveredHeld position — never continues forward blind
Communication link interruptedHeartbeat loss triggers safe state within 1.5 sControlled stop in lane, blue light pattern
Upper-layer (software) anomalySafety layer operates independently of application logicMotion output frozen by the safety boundary
Human needs controlRemote takeover and physical e-stop buttonFull mechanical stop, requires explicit re-engage
Scope statementEos supports emergency response — position sharing, alerting and scene marking. It does not diagnose conditions, provide medical treatment, or replace medical professionals. Response timelines are part of the system design; medical outcomes are not claimed.

Test, learn, improve

A pipeline, not a promise.

Each subsystem moves through the same disciplined sequence. A stage is only considered complete when its success criteria have been met and its failures documented.

01

Requirement

Written success criteria before any test begins — what must be proven, and how it will be measured.

02

Prototype

A physical build that can be measured. The current unit is a functional engineering prototype, not a render.

03

Bench test

Controlled, repeatable measurement on the bench — motors, braking, endurance, lighting.

04

Controlled field test

Closed greenway sections with defined conditions before any public use.

05

User trial

Runners, clubs and school teams use the system in supervised sessions and give structured feedback.

06

Iteration

Failures return to the top of the pipeline as new requirements. Nothing is patched silently.

Evidence trail

What has been built and measured so far.

Each card states its true status. Completed items show their activity; planned items are labelled as planned. Open a card for the full evidence drawer.

HONEST LABELSNo footage is published yet, so every media slot says "Footage scheduled" instead of showing a playable video. Measured cells stay "Pending measurement" until a real test runs with documented conditions.

Key metrics

Current results vs. next-stage targets.

The table distinguishes, row by row, between what has been measured and what is a goal. Where data does not exist yet, the cell says so.

MetricCurrent resultNext-stage targetStatus
Follow success rateControlled-field test — data collection in progress≥ 95%Test target
Mean follow-distance errorBench measurement pending≤ ±0.5 mTest target
Obstacle stop distanceBench test pending≤ 1.5 mTest target
Single-session enduranceBench test pending≥ 2.5 hTest target
Night visibility rangeDesign validation pending≥ 200 mTest target
Runner & club UX ratingInterviews underway; scoring protocol defined≥ 4.0 / 5Study in progress

METHODOLOGYEach published result will include: test procedure, environment (surface, light, weather), sample size, pass/fail criteria, raw data link and iteration notes. A row marked "Test target" contains no measured value yet — targets are engineering goals for the next stage, not results.

Development roadmap

What comes next, in order.

  1. Now

    Bench validation of motion & braking

    PWM control curves, ramp times, stop distances and endurance on the bench. First controlled-field follow runs.

  2. Phase 2

    Controlled field trials on greenways

    Closed greenway sections with obstacle and night scenarios. Success criteria published before each trial.

  3. Phase 3

    User trials with runners, clubs and schools

    Supervised sessions with structured feedback; UX scoring against the ≥ 4.0/5 target.

  4. Phase 4

    Multi-unit group coordination

    Group map, pace bands and multi-runner support in real club training.

  5. Vision

    Mobile support infrastructure

    Fleet-based support for parkruns, school programmes and marathon camps — the greenway service layer.

References

External sources used across this site.