Skip to main content

Case Study – Autonomous PV Farm Inspection

Energy & Utilities

How it works: From first sensor to fleet deployment - one platform to connect, simulate and control. Connect with WiseOS, control via MCP ROS2, validate in Digital Twins.

Summary

A large solar PV farm wanted to reduce manual inspection effort and improve visibility into the condition of panels, structures and substations.

WiseVision delivered a solution that:

  • Connected LoRaWAN sensors, inspection drones and ground robots via WiseOS
  • Used MCP ROS2 to coordinate inspection missions and data collection
  • Built a Digital Twin of the PV farm to test routes and procedures before live deployment

Stack Used

  • WiseOS
  • MCP ROS2
  • Digital Twins (O3DE / Isaac Sim / Genesis)
  • LoRaWAN
  • ROS 2

Key Benefits

  • Single telemetry and control fabric
  • Repeatable inspection missions
  • Historical logs in one place
  • Safe experimentation using Digital Twin

Challenges

PV farm spread across a large area with limited connectivity in some zones

Manual inspections were time-consuming and difficult to standardize

Sensor data, drone imagery and robot telemetry were handled in separate systems

Changes to inspection routines were hard to test without disrupting operations

Solution

1. WiseOS – Operational Layer

WiseOS was deployed as the central operating layer:

  • LoRaWAN sensors integrated via the WiseOS LoRaWAN↔ROS 2 bridge
  • Drones and ground robots connected over ROS 2 with Zenoh-based transport
  • Telemetry stored in the Data Black Box (InfluxDB)

2. MCP ROS2 – Mission Orchestration

MCP ROS2 provided tool-based access:

  • List robots, sensors and available missions
  • Pre-defined inspection workflows as MCP tools
  • Launch drones, subscribe to feeds, log events
  • Script missions without low-level glue code

3. Digital Twin – Validation

A Digital Twin using O3DE / Isaac Sim / Genesis:

  • 3D layout of panel rows and substations
  • Simulated robots and drones
  • Integration with WiseOS and MCP ROS2
  • End-to-end testing before live deployment

Architecture Overview

  1. LoRaWAN sensors → ChirpStack → WiseOS LoRaWAN bridge → ROS 2 topics
  2. Robots & drones → ROS 2 nodes → WiseOS (Zenoh transport, Data Black Box)
  3. MCP ROS2 server → exposes ROS 2 interfaces as MCP tools for operators/agents
  4. Digital Twin → subscribes to ROS 2 topics and interacts with MCP ROS2 for simulated missions

Outcomes

Illustrative outcomes for this type of deployment:

✅ Operational Efficiency

  • More consistent inspection routines with repeatable missions
  • Reduced time spent on manual walk-through inspections

✅ Better Traceability

  • Historical telemetry and mission logs in one place
  • Single unified stack for robotic inspections

✅ Risk Reduction

  • Safer experimentation using the Digital Twin before live changes
  • Test routes and procedures virtually first

✅ Scalability

  • Reusable architecture across additional PV sites or wind farms
  • Easy integration with third-party analytics

Next Steps

The same architecture can be extended to additional sites, wind farms, or integrated with advanced analytics and AI-based route planning