Industrial-grade 2D LiDAR sensors for AGV, AMR and conveyor lines now span 0.1 m to 40 m at 1–10,000 Hz frame rates with sub-2 W power budgets, and 3D units have crossed 10 million points per second on 128-channel silicon — the operating envelope most line builders are standardising on in 2026 [S3][S6].
Three vendor clusters are defining the spec window: Chinese dToF specialists such as Benewake shipping TF-series single-point and array modules from 1.35 g up to 86 g; North-American digital LiDAR houses Ouster (REV8, OS1) pushing megapixel-class 3D with native color fusion; and emerging solid-state / mechanical-hybrid builders including Richbeam with the LakiBeam, KoraBeam, LoraBeam and HaloBeam120/240 lines [S1][S2][S3]. MATLAB's Lidar Toolbox lists Velodyne, Ouster and Hokuyo as currently supported third-party hardware packages, locking the engineering-workflow side of the stack [S4].
Sensing modalities and what each one actually delivers on a plant floor
LiDAR divides cleanly into dToF (direct Time-of-Flight), FMCW and phase-shift ranging, with dToF dominating short-to-medium range industrial products in 2026 [S3]. Benewake's TF-UW500 underwater dToF unit outputs at 20 Hz on UART / I²C under 0.5 W in a 24 × 16 × 20.4 mm package, and the TF02-Pro delivers 0.1–40 m at 1–1,000 Hz for 1 W — the typical mid-range industrial 2D envelope [S3].
For 3D perception, Ouster's OS1 mid-range sensor hits 90 m at 10% reflectivity, 250 m maximum, 0.0 m minimum range, and 10.4 million points per second across 128 vertical channels, with a digital architecture that fuses color in silicon rather than post-processing [S6]. Ouster positions its REV8 line as "double the range, double the resolution, in native color" with the L4 silicon stack, marketed for harsh-environment mining and industrial automation [S2].
Solid-state LiDAR rounds out the field: Richbeam's HaloBeam120 covers a 120° FOV and HaloBeam240 a 240-line FOV, both framed as automotive-grade units for blind-spot detection and autonomous logistics, while the LakiBeam 2D scanner is Richbeam's "high-stability bestseller" for AGV and traffic monitoring [S1]. The SICK, LEUZE and KENCY ecosystem, plus China's Tianjin G-TEK Sensor Technology, continues to define the European and Chinese 2D-industrial base [S5].
Selection criteria engineers actually check before signing a PO
Five numbers drive the spec sheet in 2026: range, frame rate, point cloud density, FOV, and power consumption — and the vendor data shows they trade off predictably [S3]. A TF-Luna at 0.2–8 m, 1–250 Hz, 0.35 W and < 5 g is the typical "short-range, low-power" pick; a TF03 at 0.1–180 m, 1–9,800 Hz, IP67 and 86 g is the long-range industrial workhorse; a TFA300 reaches 0.1–290 m at 1–10,000 Hz for 0.45 W in a 10.5 g package, the densest power/weight ratio in the lineup [S3].
Interface choice matters as much as optics: UART, I²C and I/O dominate short-range boards, while CAN and RS-485 show up the moment the sensor is mounted on a moving AGV or mobile robot — TFmini-i, TF02-i and TFA300 all use CAN/RS-485 specifically for vehicle-bus integration [S3]. Protection class separates enclosure-less modules (TF-Luna, TFA170-L, TF-NOVA) from IP65 (TFmini-S / TF02-Pro) and IP67 (TF03, TF350) units, and that is usually the first line item buyers filter on.
For 3D, the OS1 spec sheet puts 90 m at 10% reflectivity as the headline figure versus 250 m absolute max, 0.0 m minimum range, and 10.4 M points/s — the working combination most automation buyers compare against [S6]. Richbeam's industrial claim is more mechanical: the LoraBeam "industrial design enables 40% smaller size, redefining cost-effectiveness compact laser sensing" — a real packaging claim on a 2D navigation unit, not a 3D range claim [S1].
Use cases: AGV, AMR, perimeter security, volume and stack-loop automation
Inside a plant, 2D LiDAR is the de-facto safety scanner for AGV and AMR — the KoraBeam obstacle-avoidance LiDAR with "regional warning function" is explicitly sold into that loop, and the LakiBeam 2D scanner is positioned for "robotics, AGVs, traffic monitoring, and industrial automation" [S1]. Benewake's TF03 at 1–9,800 Hz is the typical anti-collision choice on outdoor autonomous logistics; the TFS20-L at 21 × 15 × 7.87 mm and 1.35 g goes onto drone-class platforms where mass matters more than range [S3].
3D LiDAR is moving into perimeter security and smart infrastructure: Ouster's OS1 is described as "a mid-range sensor balancing resolution and range, making it a versatile choice for security, smart infrastructure, and autonomous vehicles" [S6]. Volume measurement is its own application — Richbeam's BeamMaster1 ("Smart 3D Laser scanner for accurate and on-demand volume data") and BeamMasterX ("Stand-up 3D laser scanner") plus the Silo Volume Monitoring case study (brewery grain silos) show LiDAR displacing load cells and ultrasonic level gauges where non-contact, on-demand reading wins [S1].
For smart camera integration on a line, LiDAR feeds the 3D point cloud into perception stacks while cameras deliver texture — Ouster's "native color" silicon fusion claim is the explicit move to collapse that two-sensor stack into one hardware path [S2]. For smart valve positioner feedback loops, single-point dToF modules like the TF02-Pro-W (0.1–25 m, IP5X, RS-485) are increasingly used for valve-travel verification on outdoor piping where ultrasonics struggle in steam, a job that traditionally went to displacement sensor probes [S3].
Functional safety, durability and the standards anchor for 2026
Ouster's REV8 line is explicitly engineered "for functional safety" and designed for "ASIL-B, SIL-2, and PLd functional-safety certifications," which is the language most European and US automotive Tier-1s now require from a 3D LiDAR vendor [S2]. The same source states that the REV8 is "the highest-performing 3D lidar family, designed for functional safety" and "cybersecure" — both claims show up in mining deployments where "shock, vibration and temperature constraints" drive sensor selection, per Matt Reiland, Technical Director, Automation Innovation at Komatsu [S2].
On the durability side, IP65/IP67 enclosures (TFmini-S, TF02-Pro, TF03, TF350) and "automotive-grade" framing for the HaloBeam120/240 are the spec lines that satisfy harsh-environment specs [S1][S3]. Solid-state framings — "Pure solid-state LiDAR" — sit on the 120°/240° FOV HaloBeam units and target blind-spot and autonomous-logistics duty, replacing legacy 2D safety scanners where wider FOV is needed without spinning optics [S1].
The MATLAB Lidar Toolbox integration (Velodyne, Ouster, Hokuyo support packages, earliest release R2020b / R2022a through "Current") is the closest thing to a de-facto engineering workflow anchor in 2026, and is what most Tier-1 perception teams cite when they say "we can simulate this sensor before we buy 50 of them" [S4].
Comparison: where each LiDAR class fits the 2026 spec window
The cleanest 2026 line-up runs: 2D short-range dToF (TF-Luna, 0.2–8 m, 0.35 W) → 2D mid-range dToF with industrial enclosure (TF02-Pro, 0.1–40 m, 1 W, IP65) → 2D long-range dToF (TF03, 0.1–180 m, 1 W, IP67) → 3D mid-range digital (Ouster OS1, 90 m @ 10%, 10.4 M pts/s, 128 channels) → automotive-grade solid-state wide-FOV (Richbeam HaloBeam240, 240° FOV) [S1][S3][S6]. On the four decision axes most specifiers weigh — range, frame rate, FOV, power — the 2D dToF class wins on power and frame rate, the 3D digital class wins on point density and color fusion, and the solid-state class wins on FOV and mechanical ruggedness [S1][S2][S3].
The 2D units that fit AGV safety loops (KoraBeam, LakiBeam, TF02-i) trade point density for low cost, deterministic scan rates and CAN/RS-485 bus integration — a TF02-i delivers 0.1–40 m, 1–1,000 Hz, ≤ 0.85 W @ 12 V, IP65 in a 60 g package with CAN / RS-485 [S1][S3]. The 3D units that fit perimeter and digital-twin loops (OS1, OS1 Max, REV8) trade cost for point density, native color and ASIL-B / SIL-2 / PLd safety framing [S2][S6].
For a 2026 line build, the practical decision is: pick 2D dToF for vehicle safety and short-loop detection, pick 3D digital for security, digital-twin capture and outdoor autonomy, and pick solid-state wide-FOV only where mechanical scanning parts are a hard no — and verify functional-safety coverage against your end-application target (ISO 13849 for AGV safety, IEC 61508 for process-industry SIL) [S2][S3][S6].
Limitations, failure modes and what the spec sheets still don't tell you
Range numbers are quoted at low-reflectivity targets (Ouster OS1: 90 m @ 10%) — a black conveyor belt at 30 m may register at 50% of the rated distance, and outdoor LiDAR at noon on snow behaves very differently from the same unit in a warehouse aisle [S6]. Most 2D units publish frame rate as a "1–N Hz" programmable range (TF03: 1–9,800 Hz; TF-NOVA: 1–900 Hz), but the effective angular resolution falls as the scan rate climbs — a 9.8 kHz TF03 scan is not the same point cloud density as a 100 Hz scan, and spec sheets do not always make that explicit [S3].
Solid-state wide-FOV LiDAR still has to prove outdoor automotive reliability against spinning mechanical units, and the HaloBeam120/240 are explicitly framed as "automotive-grade" but the broader 2D market is still dominated by mechanically scanned units from SICK, LEUZE, KENCY and G-TEK — meaning the spec sheet's "solid-state" claim has to be matched to your maintenance philosophy [S1][S5]. Finally, "native color" is a silicon-level fusion, not a software post-process — buyers expecting an RGB camera fused onto a point cloud at the host side should verify whether the Ouster REV8 output is the silicon-fused stream or a host-side composite [S2].
On the additive manufacturing material side, parts printed for LiDAR brackets and sensor mounts now commonly spec PA-CF or PA-GF for stiffness-to-weight, since a 1.35 g TFS20-L drone mount and a 10.5 g TFA300 long-range bracket are both mass-constrained, and that pushes the mechanical side of the LiDAR stack into the same digital-twin and machine vision system manufacturing workflow that other 2026 factory-sensor lines are converging on [S3].
Sourcing signals worth tracking through the rest of 2026
Three things to watch: (1) the Ouster REV8 functional-safety certification closure (ASIL-B / SIL-2 / PLd) and the first OEM design wins against it [S2]; (2) the next dToF single-point module refresh from Benewake — frame-rate ceilings and IP67 power budgets are the specs most likely to move next, with the TFA300's 10 g / 290 m / 0.45 W combination setting the bar [S3]; (3) the solid-state wide-FOV ramp from Richbeam's HaloBeam line (120° / 240°) and any automotive-grade deployment data published against it, which is where the 2D-to-3D boundary in industrial automation is most likely to break first [S1]. machine vision in smart manufacturing rollouts will pull the LiDAR spec with them, and that is the integration layer to monitor in the second half of 2026.