/ Glossary /
AI & Robotics, in plain English.
Search the terms that show up across our platforms, case studies and blog posts.
/ Robotics /
LiDARLight Detection and Ranging: a sensor that measures distance by timing laser pulses reflected off surrounding surfaces, producing precise 3D spatial data for navigation and mapping. Used across Leafloat's Howar, Sadeem, and Robotic PC platforms for GPS-denied navigation and site surveying.PTZ CameraPan-Tilt-Zoom camera: a remotely steerable camera with optical zoom, used for live remote site monitoring and detailed visual inspection from a safe distance. Deployed on Howar and in Construction industry site-monitoring workflows.
/ AI & ML /
Bounding BoxesRectangular markers a computer vision model draws around detected objects in an image or video frame, used to localise and classify items such as missing PPE, defects, or vehicles. A core building block of the vision AI models behind Amanex AI and Howar.Vision AIComputer vision models that interpret camera or sensor imagery to detect, classify, or measure real-world conditions — from PPE compliance to structural cracks — without a human reviewing every frame.
/ Autonomy /
GPS-Denied EnvironmentAny indoor, underground, dense-canopy, or heavily obstructed space where standard satellite positioning is unreliable or unavailable, requiring LiDAR, IMU, or visual-inertial navigation instead.Point CloudA dataset of thousands to millions of 3D coordinate points captured by LiDAR or photogrammetry, representing the precise shape of a physical space or object. Sadeem processes point cloud data in-browser into measurable 3D models.
/ Manufacturing /
Autoclave Prepreg CuringA composite manufacturing process where pre-impregnated carbon fibre sheets are cured under heat and pressure inside a pressurised oven (autoclave), producing parts with maximum density and zero voids.CNC TrimmingComputer Numerical Control trimming: automated, programmable routing used to cut composite or metal parts to exact tolerances after moulding or layup, removing manual variation from the finishing process.Predictive MaintenanceUsing sensor data and machine learning to forecast equipment failure before it happens — typically 48–72 hours in advance — rather than relying on fixed maintenance schedules or reacting after a breakdown.
/ Construction /