Glossary / Core technology
What is Simultaneous Localisation and Mapping (SLAM)?
SLAM is the technique that lets a robot build a map of an unfamiliar space while simultaneously working out its own position within that map.
Last reviewed:
SLAM is the software layer that turns raw LiDAR and camera data into something a robot can actually use to navigate. As the robot moves, it continuously compares new sensor readings against the map it's building, refining both the map and its own estimated position at the same time — hence 'simultaneous'.
This is what allows a robot to be walked or driven around a building once during setup and then operate independently afterwards, rather than needing GPS, which doesn't work reliably indoors, or fixed markers, which AGVs require. Good SLAM implementations cope with dynamic changes — furniture moved, doors closed — without losing track of where they are.
SLAM quality is one of the biggest differentiators between robot brands. Poor implementations drift over time, lose their position in featureless corridors, or fail to recover after a knock. This is worth asking a supplier about directly rather than taking on trust.
Common questions
- How long does SLAM mapping take for a new site?
- Initial mapping typically takes an hour or two for a mid-sized building, done once during setup and updated when layouts change significantly.
- Can SLAM cope with a building that changes layout often?
- Modern SLAM handles day-to-day changes like moved furniture well; major layout changes such as a refit usually need a re-map.
- Does SLAM need Wi-Fi or GPS?
- No, SLAM works using onboard sensors alone, which is why it's suited to indoor spaces where GPS doesn't reach.
Related terms
Next steps