EIRS

All demos · Industrial challenges

Robot telepresence · Remote inspection · Immersive robotics

Immersive Telepresence for Navigation and Inspection

An operator can guide a mobile robot through places that are difficult or unsafe to access and view its live camera feed in virtual reality (VR). Images tied to locations in the scene support later specialist review. The framework connects live robot control, spatial context and detailed follow-up, with opportunities to adapt it to other robots and capture equipment.

Immersive Telepresence for Navigation and Inspection
Robot navigation with camera views and the immersive operator in view.

Why remote inspection matters

Some environments are difficult or unsafe for people to access. This can include industrial spaces that are hard to reach, hazardous or accident prone areas, remote infrastructure and places beyond practical reach. Mobile robots can carry cameras into these spaces and give operators a live view while they guide movement. Captured views and marked points of interest help specialists examine details after the live inspection.

What it enables

  • Assess hard-to-reach areas remotely before sending people into them.
  • Guide a mobile robot while monitoring what its cameras can see.
  • Change viewpoints to inspect details and understand the surrounding space.
  • Capture important images and points of interest for specialist review.
  • Evaluate a remote inspection workflow before adapting it to a particular site or robot fleet.

Testbed and equipment

The current testbed connects a Unitree AlienGo quadruped robot and its onboard cameras to a Unity VR application. Operators use a Meta Quest Pro headset with an Omnideck VR omnidirectional treadmill or a joystick for movement. Head movement or joystick input can change the camera view. A Matterport LiDAR scan, a laser-scanned spatial model, provides the virtual scene for navigation and spatial context.

How the inspection framework works

The framework links the operator interface, robot communication and inspection review. A Matterport LiDAR scan is imported into Unity to provide a spatial model. Live camera video gives the operator a view from the robot, while the reconstructed environment provides spatial context. Captured scene information can also be explored later for detailed review.

Three layers built around ROS, the Robot Operating System, connect the robot and immersive application. Robot nodes handle sensing and movement. Middleware translates device data and commands. Containerised services in a Docker-based private cloud coordinate communication between components. Movement commands travel from the VR application to the robot, while live video and scene updates return to the operator.

The operator can save geotagged screenshots, which attach an observation to a location in the scene. These points of interest can be shared through a central platform so a specialist can review selected details and recommend follow-up.

How operators use it

The operator guides movement with the Omnideck or a joystick and changes the camera view using head movement or joystick input. Live video supports inspection during navigation. When a detail needs further attention, the operator captures an image linked to its scene location. A specialist can then examine that observation alongside the reconstructed environment.

Where this could apply

Teams can use this approach to assess areas that are hard or unsafe to reach, guide a robot toward points that need closer inspection and share selected images with specialists. It could support inspections in restricted process spaces, remote infrastructure, accident prone locations and emergency response. A site adaptation can align robot controls, camera views and evidence handover with the inspection workflow.

Next development opportunities

Further development could connect additional ground robots and drones, add camera and sensing options such as thermal imaging and LiDAR, and support multiple live viewpoints and shared specialist review. Site adaptations could also align the control interface, safety procedures and evidence handover with each team’s needs.

Industrial applications

Remote inspection involves more than getting a camera into a difficult location. Operators need to understand their surroundings, direct movement, inspect details and communicate useful observations to others.

Immersive robot teleoperation

Develop remote controls and immersive views around the robot’s task. A project could evaluate how movement inputs and feedback help operators guide the equipment.

Inspect difficult environments remotely

Combine live exploration with detailed review of captured environments. The approach could be adapted to locations where access is difficult, hazardous or impractical.

Design remote camera views around the task

Connect camera viewpoints and controls with what operators need to inspect. A project could compare ways of directing the view while maintaining awareness of the surrounding space.

Capture and share inspection observations

Associate important images and points of interest with their location in the inspected environment. This can provide a foundation for specialist review, reporting and follow-up.

Develop a workflow around your site and robot

A collaboration could connect the relevant robot, cameras and inspection procedures, then evaluate the complete workflow from remote operation to specialist review.

Explore research and collaboration →

Media

Videos

Images

Framework overview connecting VR control, robot cameras and specialist review.
Framework overview connecting VR control, robot cameras and specialist review.
Mobile robot fitted with camera and sensing equipment.
Mobile robot fitted with camera and sensing equipment.
Camera and computing equipment mounted on the mobile robot.
Camera and computing equipment mounted on the mobile robot.
An operator uses an Omnideck VR omnidirectional treadmill to guide a mobile robot.
An operator uses an Omnideck VR omnidirectional treadmill to guide a mobile robot.
Robot movement and camera view controls for immersive telepresence.
Robot movement and camera view controls for immersive telepresence.
VR view of a reconstructed tunnel with the operator in view.
VR view of a reconstructed tunnel with the operator in view.
A close view of tunnel geometry and spatial markers in VR.
A close view of tunnel geometry and spatial markers in VR.
Immersive view of a tunnel junction and its pipework.
Immersive view of a tunnel junction and its pipework.
A mobile robot and VR operator share the Omnideck VR omnidirectional treadmill setup.
A mobile robot and VR operator share the Omnideck VR omnidirectional treadmill setup.

Research and reports

Paper
Remotely Operated Robots for Underground Applications
Berg Huettenmaenn Monatsh · 2025

Project and programme

This work is part of DDIA, Data Driven Immersive Analytics in Digital Industries.

DDIA is an FFG COMET Module (2022 - 2026) at Know Center Research GmbH.

↑ Top