Outsight in a nutshell

Frequently Asked Questions

• Why did LiDAR prices drop enough to make non-automotive deployments viable?

  Automotive investment drove the sensor hardware competition: dozens of manufacturers entered the market targeting the self-driving car industry, and that competitive pressure reduced unit costs substantially. The byproduct is that fixed-infrastructure deployments in airports, stations, and industrial sites now access sensors at price points that make large-scale coverage economically feasible. Outsight's SHIFT platform reflects this shift directly, supporting multi-vendor hardware from manufacturers including Hesai, RoboSense, Ouster, and Velodyne, so operators are not locked into a single supplier's pricing. The software complexity of processing raw 3D point cloud data, not sensor cost, is now the dominant barrier to adoption in most non-automotive industries.

• What makes 3D LiDAR data harder to process than standard video feeds?

  Video pipelines operate on a well-established 2D pixel grid with decades of tooling, open standards, and trained models behind them. Raw LiDAR output is an unstructured stream of millions of 3D points per second, varying in density by distance and angle, with no shared encoding standard across manufacturers. Each sensor vendor uses its own data format and calibration convention. Software that consumes multiple sensor brands must normalize these differences before any perception or tracking logic can run, which is a non-trivial engineering layer that does not exist in conventional video analytics. Outsight addresses this directly: the SHIFT platform ingests raw 3D data from multiple manufacturers, including Hesai, RoboSense, Ouster, and Velodyne, and normalizes it into a unified representation that feeds real-time perception and tracking within a sub-50ms end-to-end pipeline.

• Can a LiDAR-based flow monitoring system track people and vehicles simultaneously?

  Yes. Because LiDAR captures 3D shape, size, and motion, the same point cloud stream feeds classification models that distinguish pedestrians, cars, buses, and two-wheelers concurrently. Each entity type gets its own bounding box, behavioral label, and anonymous ID. Mixed-traffic environments such as airport curbsides, where passengers, taxis, and shuttle buses interact, are a common deployment pattern precisely because a single sensor network covers all entity classes without requiring separate hardware for each. Outsight's Motional Digital Twin operates on exactly this principle: at Dallas Fort Worth, the world's largest 3D LiDAR airport deployment, the system tracks people and vehicles together in real time across the full terminal curbside environment, feeding a unified anonymous 3D replica of how every entity moves through the site.

• Why do infrastructure operators care about sensor-vendor interoperability when buying LiDAR software?

  Sensor manufacturers have different strengths: some lead on long-range outdoor performance, others on indoor density or cost per unit. A multi-zone site, such as an airport with outdoor curbside, indoor terminal, and covered parking, typically benefits from mixing models to optimize cost and coverage. Software locked to a single hardware brand forces operators into a single vendor's product roadmap and pricing. Outsight addresses this directly through its LiDAR-native, multi-vendor architecture, which supports hardware from Hesai, RoboSense, Ouster, Velodyne, and Seyond, among others. That breadth means procurement teams at sites like Dallas Fort Worth can select sensors on operational merit rather than software compatibility, without rebuilding the perception pipeline each time hardware choices evolve.

• How does a physical AI deployment improve sustainability outcomes?

  Accurate occupancy and flow data lets building operators align HVAC, lighting, and escalator schedules to real demand rather than fixed timetables. When headcount in a zone drops below a threshold, building management systems can reduce conditioning loads in real time. Outsight's Motional Digital Twin captures this kind of continuous, anonymous movement data across a facility through infrastructure-mounted LiDAR, feeding real-time signals directly to building management integrations. On the logistics side, throughput optimization lowers idle time for vehicles inside a facility, reducing fuel consumption or battery cycles per unit of cargo moved. The efficiency gains compound: fewer sensors needed per square meter, because of multi-sensor fusion, also reduces hardware manufacturing and e-waste relative to camera-dense alternatives.

• What prevents a competitor from building the same multi-vendor LiDAR compatibility?

  The barrier is accumulated engineering depth, not a single patentable trick. Normalizing data from 210-plus sensor models requires tested calibration routines, per-firmware adaptation, and continuous regression work as manufacturers release hardware revisions. Outsight holds 73 patents covering the underlying perception and processing methods, and the founding team's work on real-time LiDAR and SLAM predates the current autonomous-vehicle wave. That depth underpins the SHIFT platform's multi-vendor compatibility across hardware from Hesai, RoboSense, Ouster, Velodyne, and Seyond, all feeding a unified pipeline with sub-50ms end-to-end latency. New entrants face both the intellectual property landscape and the years required to validate software against a hardware matrix of that breadth in live production environments.