Robotics glossary
Robotics glossary
Plain-English definitions for the perception terms you'll see across our site and spec sheets.
- Perception
- The sensors and software that let a robot sense its environment — LiDAR, cameras, depth, safety fields, and the compute that fuses them.
- Point cloud
- A 3D set of points measured by a LiDAR or depth sensor, each with x/y/z (and often intensity). The raw geometry a robot uses to see.
- SLAM
- Simultaneous Localization and Mapping — building a map while tracking where the robot is inside it.
- Localization
- Knowing where the robot is in a map or facility frame.
- Pose
- Position + orientation of the robot (or sensor) in space.
- Field of view (FOV)
- The angular region a sensor can see. 360° LiDAR sees all around; a forward camera sees a cone.
- Obstacle avoidance
- Detecting and reacting to obstacles — slowing, replanning, or stopping.
- Safety-rated sensor
- A sensor certified for functional safety stop functions (e.g. Type 3 / SIL2 / PL d).
- Protective field
- The safety zone where a person or object triggers a stop.
- dToF
- Direct time-of-flight depth — measures distance per pixel by timing light pulses.
- ROS / ROS 2
- Robot Operating System — the common middleware stack for robotics software.
- Edge compute
- On-vehicle processing that runs perception and inference without relying on the cloud.
Ready when you are
Ready to get specific?
Bring your application and we'll translate it into the right sensing stack.