Direct ingestion
Authorized MQTT ingestion through AWS IoT Core, with scoped device identity and routing.
Planned · V2RUPURUPU / OPERATIONS
Fleet overview.
Your fleet at a glance. A clear path to what needs attention.
MONDAY, 28 SEPTEMBER 2026Sample snapshot · 11:30 EAT
One source needs your attention
Review source ↗Naivasha mirror last synced 47 minutes ago. Device status may be out of date.
Device health
Latest available observations from approved sources.
| DEVICE / ID | SITE | HEALTH | BATTERY | OBSERVED AT ⓘ | Details |
|---|
Six classes. One fleet.
Commissioned devices by class
HealthyAttentionStale
Source freshness
Read-only client mirrors
⇄
Nanyuki mirrorSynced 11:28 EAT
Fresh⇄
Naivasha mirrorSynced 10:43 EAT
Stale⇄
Thika mirrorSynced 11:27 EAT
FreshSource freshness is separate from device connectivity.
Naivasha source is stale
The sample freshness threshold is 15 minutes. Last synchronized at 10:43 EAT.
A defined data boundary
Illustrative access policy for this V1 workspace.
Visible in Precision
Device identity, class, firmware, battery, connectivity, fault codes and source timestamps.
Stays with the client
Primary measurements, farmer context, scores and farmer communications.
Requires separate agreement
Retention, field permissions, commands, firmware updates and notification authority.
This concept uses sample records. No client databases are connected and no live data is stored.
Gateway management
Versioned components on qualified OS gateways. Canary rollout, health verification and tested rollback.
Planned · V2Client-defined routing
Shanga-hosted, Shanga-relay or client-first routing, with independent storage and access policies.
Planned · V2Qualification comes first
Gateway management and sensor firmware are separate workflows.
Compatible gateways
Greengrass needs a qualified OS gateway. ESP32-based Heltec devices require a Linux companion or replacement.
Signed sensor updates
MCU firmware updates require signed artifacts, device-side verification and a tested recovery path.
Independent local alarms
Critical local alarms remain independent of cloud availability. Cloud status resumes after reconnection.