• •
T
🖥️
Teleste

DEMO
·
%
critical ·

DEMO
Coords
🌐 IP
Coords
served board

Status
–

· ·

DEMO


– · ·

🔴
🟡
✓
✓

–

·
·
·

·
·
→

·

, , .

DEMO
— none —
— none —

No displays known to the data plane yet — a PIDS appears once it pulls content (or after ).

: : ⚠

⚠

Associate a display with a live SIRI vehicle. The display then rides that vehicle on the map (feed → vehicle → live, like the GTFS chain). Or bulk-assign via CSV: export → fill the column → re-upload.

Teleste TFT/RGB displays managed by an IQU HUB. HUB injects content via corelink; map each display to its IO-Core stop here. Configure the connection in .

DisplayOS displays (config-set owned)

DEMO
How this is calculated

The display wakes once a minute and runs one cycle, so a day is 1440 cycles. Every cycle is assigned to exactly one of three measured profiles, and the consumption of the day is the sum of those cycles plus the passenger interactions.

P1 runs when the next departure is within one minute. The display rewrites the whole page because a new stop replaces the current one, and the modem connects to the server for line information. It is the most expensive profile.

P2 runs when the next departure is between one and two minutes away, and once every hour when the device pushes its reports to the server. The modem is on; the display updates only part of the page.

P3 runs the rest of the time — the clock changes and nothing else. The modem is off. In a typical stop this is most of the day.

How the day is divided. Outside the service window there are no departures, so every cycle is P3 apart from one P2 per hour for the report push. During rush hours a departure arrives every minute, so every cycle is P1, which absorbs both the P2 window and the push. Rush-hour departures are taken from the daily departure figure, never added to it: 60 departures per rush hour, so the model applies at most departures ÷ 60 rush hours and says so under the input when it limits them. This is also why more rush hours can lengthen the estimated life: the same departures, packed one per minute, need no separate P2 cycle each — that is what the firmware measures, not an extra cost. For the rest of the service window the split depends on how close together the departures are: each departure claims one P1 cycle and one P2 cycle, and whatever is left over is P3. Once departures are less than two minutes apart there is no P3 left, and further departures start converting P2 cycles into P1.

From cycles to years. Each measured curve is integrated to give the consumption of one cycle or one button press. Those are multiplied by how often they happen, summed over the day, and divided into the usable capacity of the pack. Usable capacity is the rated capacity multiplied by the usable percentage — it is not the full nameplate figure, because a pack is not discharged to zero.

What this does not model. Temperature: cold cells deliver less capacity than the rating. Ageing: capacity falls over the service life. Signal quality: a weak network makes P1 considerably more expensive than the bench measurement. Instrument resolution: currents were measured to the nearest milliamp, so a sleep floor below that is invisible and is added separately if you enter one.

Firmware 63 and later — three-profile model. Devices on firmware 62 and earlier use the legacy two-profile model and will show a different figure for the same stop.

No canvas layouts.
Click to create one.

↗

:

render

Set your Azure Speech API key and region in . You need an Azure Cognitive Services Speech resource.
pip install piper-tts && python3 -m piper.download_voices it_IT-paola-medium --data-dir

No audio yet — generate above or upload in Content → Media → Audio.

 

→

selected
Device Stop / Location Problem Severity SLA Status Opened Closed

Maintenance to schedule before failure — battery replace-by + recurring faults. Schedule → planned ticket in JSM.

selected
Device Trigger Predicted SLA

selected

Settings

External data adapters (ports & adapters). Credentials are encrypted at rest and shown masked. Transit feeds (SIRI, MQTT, OJP) are configured in their dedicated Transit Feeds hub.

Configure the SMTP server your tenant uses to send notification emails. These settings are scoped to your tenant only.

Send test email

Configure Azure Cognitive Services TTS for your tenant. Leave key blank to use the global CMS key (if configured by superadmin).

Once configured, go to Audio / TTS in the sidebar to generate voice files.

Export all data and files for your tenant only into a downloadable archive. Restoring overwrites your tenant's current data. Other tenants are never affected.

💡

Danger zone

Permanently deletes all data for this tenant (devices, profiles, layouts, content, automations…). Your admin account and a safety backup are preserved — a backup is taken automatically first.

Perry is your operational copilot. It uses your own Anthropic API key — usage is billed to your account. The key is encrypted at rest and never shown in full.

The key is test-called on save; an invalid key is rejected.

Once enabled, the Perry mascot appears at the bottom-right of every page.

✓
Send test email

Rows stored
Table size
Loading stats…

Snapshots outside this window are deleted automatically every 24h.

Immediately deletes rows outside the retention window.

💡

💡

💡

Export or restore the data of a single tenant, independent of the full platform backup above.

💡

Danger zone

Permanently deletes all data for (devices, profiles, layouts, content…). Tenant admins and a safety backup are preserved — a backup is taken automatically first.

Danger zone — Full system reset

Wipes the data of every tenant on this platform. Tenant admins, system settings, audit logs and backups are preserved. A full backup is taken automatically before anything is deleted.

Technical measures in place to support GDPR compliance. Full legal compliance also requires a DPO review and a DPA between Teleste and each tenant.

✓ Tenant isolation — enforced at the database (Row-Level Security)
✓ Data minimisation — superadmin access to tenant data is gated
✓ Right to erasure (art. 17) — per-user erase in User Management
✓ Data portability (art. 20) — per-user export in User Management
✓ Audit retention — logs auto-purged after the configured period
✓ Break-glass access — time-boxed, reason-logged tenant access
Break-glass — request data access to a tenant

By default the superadmin works on the platform plane. Use this to request a time-boxed, audited window into one tenant's data — a reason is mandatory and recorded.

Audit log retention is configured above (Telemetry tab → Audit retention). All accesses appear in the Audit Trail.

Fallback email for superadmin password reset via OTP, when the superadmin account has no personal recovery email set. Keep this in a safe place — it provides full account recovery.

Platform/global key — used by Automations (AI email generation) and the Perry copilot for superadmin. Tenants set their own key under Settings → AI / Perry. Model: claude-haiku-4-5-20251001 (fast, cost-effective).

Tenants can override with their own key in Settings → Audio/TTS.

Users can sign in with their Microsoft corporate account. The first time they log in, a CMS account is created automatically with the default role below.

App Registration
Role Assignment

Map Entra group Object IDs to CMS roles. Highest matching role wins. Leave empty to use the default role above.

Test

Click the button below to open the Microsoft login page and verify the configuration is working.

Open Microsoft Login →

:
:
…
·

🖼
Click an image to select it · Images are managed in
📺
served board

🎫

✓
…

🎯

🚨

📡

→
🖼️
Perry copilot-1
›