Rajant
Control, configure and test the Rajant BreadCrumb mesh integration end to end — live BCAPI status and discovery, connection config, GPS/GNSS-denied positioning output, a conformance test suite, and a read-only BCAPI terminal. The mesh + ranging in the positioning tests are simulated; the solver (XGN/XRPN), discovery, topology and terminal are the real deployed services.
The lab observes the mesh over the BCAPI VIEW (read-only) role. INACTIVE is the honest state with no BreadCrumb hardware/credentials — nothing is fabricated. See the Config card to activate.
To bring a real BreadCrumb online, provision the VIEW-role credential secret and set rajant.enabled=true via helm — no code or image change. The password is mounted as a file, never entered here.
kubectl create secret generic xaitel-xrip-rajant-cred -n xaitel-xpf \ --from-literal=password='<BCAPI_VIEW_PASSWORD>' helm upgrade xaitel-xrip charts/xaitel-xrip -n xaitel-xpf --reuse-values \ --set rajant.enabled=true \ --set rajant.bcapiEndpoint='<BREADCRUMB_IP>:2300' \ --set rajant.credSecretName=xaitel-xrip-rajant-cred kubectl rollout restart deploy/xaitel-xrip -n xaitel-xpf
Full runbook: docs/DEPLOY_RAJANT_ACTIVATION.md(prereqs, verify, rollback, troubleshooting)
Positioning output
Position a platform inside a Rajant BreadCrumb mesh with no GNSS. The mesh discovers ranging references; one BreadCrumb is GPS-georeferenced (pins the local frame to WGS84); the platform solves its position and the relative fix becomes absolute.
The BreadCrumb mesh + ranging are simulated; the multilateration, georeferencing and relative→absolute transition are the real XGN/XRPN engine. Field sub-metre requires precise ranging (UWB / fine RTT) — RSSI alone is coarse.
| Scenario | Expect | State | Horiz err | Budget | Result | |
|---|---|---|---|---|---|---|
Exact ranges (solver fidelity) 7 nodes, georeferenced, zero ranging noise | ABS | — | — | 0.1 m | — | |
UWB / fine RTT (0.10 m) georeferenced mesh, precise ranging → sub-metre | ABS | — | — | 1 m | — | |
Typical RF ranging (0.30 m) georeferenced mesh, moderate ranging noise | ABS | — | — | 1.5 m | — | |
Coarse / RSSI-grade (0.75 m) RSSI-derived ranges → metres, not sub-metre | ABS | — | — | 3 m | — | |
No GPS anchor (relative-only) mesh present, no georeferenced node → stays relative | REL | — | — | — | — | |
Sparse mesh (4 nodes) minimum anchors for a 3-D fix | ABS | — | — | 2 m | — |
Field sub-metre needs precise ranging (UWB / 802.11mc-FTM); RSSI alone is coarse — the RSSI scenario is expected to land at metres, not sub-metre.
Rajant BCAPI test console (VIEW, read-only). Type 'help'.
Governed console: only whitelisted VIEW-role reads run (no mesh mutation). With no BreadCrumb connected the commands return XRIP's honest blocked responses.