# Deploy & Operate

The **Deploy & Operate** phase is where a completed Robora build goes from “box” into “action.” Here’s what happens step-by-step once a robot or IoT device leaves the assembly line:

1. **Onboarding via the dApp**
   * **Device Registration:** The owner scans a QR code (or enters a device ID) in the Robora mobile or web dApp. This links the physical unit’s hardware PKI key to their blockchain identity.
   * **Firmware & Agent Assignment:** The dApp shows a menu of available Synths (vision, navigation, inspection, analytics, etc.). The owner selects which Synths to deploy.
2. **Network Handshake & Authentication**
   * **Mutual TLS / PKI Verification:** The robot initiates a secure handshake with Robora’s edge-and-cloud gateway. Both sides verify their certificates (device vs. platform), preventing any man-in-the-middle or unauthorized access.
3. **Compute Job Submission**
   * **Task Packaging:** Sensor data (camera feeds, LiDAR scans, telemetry) is packaged into discrete compute jobs. For example, every 500 ms the device may batch up a 3D point-cloud chunk for SLAM mapping.
   * **Smart-Contract Escrow:** The owner’s $RBR tokens are staked in escrow against each compute job. Node operators see the incoming job request and its attached payment offer.
4. **Real-Time Inference & Control Loops**
   * **Edge Inference:** Low-latency tasks (object detection, obstacle avoidance) run on Coral TPUs co-located with the robot. Results (e.g., bounding boxes, safe-path vectors) are streamed back over a lightweight gRPC channel.
   * **Cloud Training / Orchestration:** Longer-running jobs (multi-robot coordination, large-model retraining) are dispatched to cloud GPUs. When complete, updated weights or trajectory plans are sent back and automatically integrated into the device’s local firmware.
5. **Monitoring & Feedback**
   * **Live Dashboard:** In the dApp, the owner monitors CPU/GPU usage, agent performance metrics (latency, accuracy), and token-spend in real time. Alerts trigger if SLAM drift exceeds thresholds or network latency spikes.
   * **Auto-Scaling & Redeployment:** If compute demand spikes (e.g., more robots are activated), the system can automatically spin up additional GPU nodes or switch more jobs to edge devices, all governed by on-chain service-level policies.
6. **On-Chain Logging & Settlement**
   * **Proof-of-Task Receipts:** Each compute segment produces a cryptographic receipt that’s hashed on-chain. This proves the work was completed, unlocking token disbursement to the node operator.
   * **Transparent Billing:** The owner sees a breakdown of edge vs. cloud costs, with charges settled automatically from their wallet. Historical logs let them audit every inference or training cycle.
7. **Over-the-Air Updates & Maintenance**
   * **Secure OTA Pipeline:** New agent versions, firmware patches, or security fixes are published via the same smart-contract–managed pipeline, signed with Robora’s hardware PKI and verified on-device before installation.
   * **Health Checks & Diagnostics:** Periodic self-tests (battery status, sensor calibration) run locally, with results anchored on-chain so maintenance teams can pre-emptively address wear-and-tear.

***

By threading together on-chain identity, token-escrow, decentralized compute discovery, and secure OTA updates, the **Deploy & Operate** stage ensures each Robora unit stays connected, up-to-date, and optimally performing, without manual infrastructure management on the user’s part.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.robora.xyz/technical-framework/3d-drag-and-drop-builder/deploy-and-operate.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.