Technology
From sensing to useful action.
Our approach connects robotic mobility, task-specific sensing and operational software around a defined job.
Platforms
Which robot does which job.
Illustrative platform configurations. Equipment is selected for each project.
Ground inspection robot
Wheels, tracks or legs, selected for the environment.
- Proposed activity
- Routine observation rounds in depots, plants and technical rooms.
- Example sensors
- Visual and thermal imaging, acoustic sensing, spatial sensing.
- Cortex contribution
- Route and checkpoint definition, evidence capture and links to asset records.
Aerial inspection platform
Suitable for observations of elevated or distributed assets.
- Proposed activity
- Observations of roofs, structures, lines and large sites from above.
- Example sensors
- Visual and thermal imaging, spatial sensing.
- Cortex contribution
- What to observe, evidence review and comparison over time.
Confined-space inspection platform
Suitable equipment for pipes, tanks or underwater assets, depending on the task.
- Proposed activity
- Inspection inside pipes and tanks, or of submerged structures.
- Example sensors
- Visual imaging and spatial sensing; other sensors depending on the equipment.
- Cortex contribution
- Evidence organised by asset and location, and the review workflow.
Sensing
Each sensor answers one kind of question.
Which of them a project uses depends on the task and on what can be validated on the site.
- Visual imaging
Images of the condition of an asset, a label, a gauge or an indicator.
- Thermal imaging
Surface temperature differences that may point to a condition worth checking.
- Acoustic sensing
Sound from equipment and its surroundings, such as a change in operating noise.
- Vibration sensing
Vibration measured by suitable instrumentation on or near the equipment, where it is available. A microphone does not replace it.
- Spatial sensing
The position and shape of the surroundings, so that observations can be placed and compared.
Proposed system architecture
From the point observed to the next round.
- Platform and sensors reach the checkpoint.
- The capture is stored with its time, position and sensor.
- The observation is associated with the asset it belongs to.
- A qualified person reviews what was flagged.
- Maintenance acts on confirmed findings.
- A reinspection shows whether the condition changed, and the next round starts from that result.
Cortex Prime products
Integration opportunities across Cortex Prime
These are proposed integration paths, scoped and validated per project.
KraftSync Transit
Bringing depot inspection findings into the context of fleet and charging operations.
About KraftSync TransitCityTwin
Spatial context and asset identity for observations made across a city.
About CityTwinTerraScan
Directing local checks to areas where territorial risk signals change.
About TerraScanPHAROS
Maritime infrastructure data, as a research direction.
About PHAROS
Research horizon
What we explore next.
Simulation, world models and Physical AI form part of the wider Cortex Prime research programme. They guide what we explore next; they do not imply validated autonomy across every environment.
The research programmeTechnology
Start with the task. Build the right system.
We start from the task and the site, then choose the platform, the sensors and the software that fit them.