Skip to main content

Upload a LiDAR scan or a CAD design. Get a road safety audit.

State‑of‑the‑art AI audits every mile of your LiDAR scan and connects the hazards it finds to your crash records, so you fix what puts drivers at risk first, without putting a crew on the road.

LiDAR of a road you already own, or the CAD model of one you are still designing: Bentley MicroStation and OpenRoads, Autodesk Civil 3D. How a design run works

From RDV Systems, trusted by 24 state DOTs since 2005.

Sight distance checked on a LiDAR scan of an existing road

Every station, both directions

Ten analyses, one scan

Choose what each corridor needs

  1. Ten analyses, one scan

    Choose what each corridor needs

    SSDAnalysis 1 of 10

    Stopping sight distance

    Tests at every station whether a driver can see an object in the lane in time to stop, stepping through the AASHTO speed table to the highest speed the view supports.

    What you getWhere drivers can't stop in time, what blocks the view (terrain, vegetation, a crest, guardrail, pole or sign) and a crash-risk score for ranking the fixes.

  2. PSDAnalysis 2 of 10

    Passing sight distance

    Compares the passing zones marked on the road with the sight distance a pass needs under AASHTO and MUTCD criteria.

    What you getZones striped for passing where a driver can't see far enough to pass safely, the setting for head-on crashes.

  3. ISDAnalysis 3 of 10

    Intersection sight distance

    Checks each maneuver at each intersection against the AASHTO sight triangles for a vehicle waiting to turn or cross.

    What you getWhich movements at which intersections have a blocked view, so clearing, signal or geometry work goes where it counts.

  4. TSAAnalysis 4 of 10

    Traffic signs

    Detects signs in vehicle camera imagery, reads each one's MUTCD code and size, and places it with the LiDAR.

    What you getA georeferenced sign inventory for the corridor, without a crew walking it.

  5. FSAAnalysis 5 of 10

    Flat spots

    Measures the longitudinal grade and cross slope of the pavement at every station.

    What you getThe flat spots where water stays on the road, so drainage fixes target hydroplaning and ice risk.

  6. HCAAnalysis 6 of 10

    Horizontal curves

    Measures each curve's radius and superelevation from the scan and checks them against the AASHTO Green Book, with maximum superelevation set per route.

    What you getCurves whose safe speed falls below the posted speed, the candidates for advisory speeds, chevrons or re-banking.

  7. GLRAnalysis 7 of 10

    Sun glare

    Builds a year-long calendar for every station of the times the sun sits near the driver's line of sight.

    What you getWhich stretches blind drivers, and at what times of day and year.

  8. GRWAnalysis 8 of 10

    Guardrail warrant

    Evaluates embankment height and fill slope against the barrier warrant in the AASHTO Roadside Design Guide.

    What you getWhere a barrier is warranted along the corridor, and where one is not, before crews install or remove guardrail.

  9. CS3DAnalysis 9 of 10

    Cross-Section 3D

    Cuts 3D cross-sections through the embankments along the corridor.

    What you getSide-slope geometry for designers, delivered as DXF cross-sections without sending a survey crew.

  10. CZAnalysis 10 of 10

    Clear zone

    Measures the clear zone width and side slopes, and finds the fixed objects and guardrails inside it.

    What you getTrees, poles and slopes in the recovery area, where a vehicle that leaves the road has no room to recover.

    Request a demo

    The corridor is simulated. Every check on it, and every number in the readout, is the analysis's own arithmetic on that geometry, against AASHTO and MUTCD criteria.

Know where your network puts drivers at risk, and what to fix first

123BIMLive turns a LiDAR scan you already own into a safety audit of the whole corridor: what is deficient, what is blocking the view, and which fixes buy the most safety per dollar.

  • Audit every mile, not a sample

    A field audit covers the sites someone can be paid to visit. Here every station on the corridor is checked, in both directions, so nothing ends up low priority only because no one drove it.

  • Spend the safety budget where it prevents crashes

    Findings come ranked by crash risk. The Safety Dashboard tests them against your state's crash records and prices the fix — crashes prevented, annual benefit and benefit-cost ratio, in the form an HSIP application asks for.

  • Nobody stands in live traffic

    Nothing is measured from the shoulder. No lane closures, no traffic control, no crew exposed to the road you are auditing.

  • Findings that hold up in review

    Every check follows the AASHTO Green Book, the Roadside Design Guide and the MUTCD, applied the same way at every station, with the driver's view from the scan behind each verdict.

  • One upload, every analysis

    Sight distance, curves, clear zone, glare, guardrail and the rest all run on the same scan. Adding one later reuses that scan instead of sending anyone back out to collect the data again.

  • Something to show, not just a report

    Drive-through video, an online report for every road and a 3D viewer in the browser — what a commission meeting, a public hearing or a funding request actually needs.

From point cloud to findings, without a workstation

One upload runs the whole pipeline in the cloud, on hundreds of parallel jobs.

  1. Upload the scan

    LAS and LAZ point clouds, multi-gigabyte files included. Uploads resume after a dropped connection, and scans in state plane feet or WGS84 are reprojected to the project's coordinate system automatically.

  2. Prepare the point cloud

    The scan is split into tiles, the ground is classified, a Point Transformer V3 model finds the road surface and parked and passing vehicles are removed.

  3. Run the analyses

    Sight lines are rendered on GPUs station by station, across hundreds of parallel cloud jobs, and an AI model names what blocks each failing view.

  4. Deliver the findings

    Reports, an online report for every road, drive-through video, CSV tables and shapefiles land in the project folder, ready to review.

One run, the whole package

Everything lands in the project folder together — nothing to request, export or chase afterwards.

  • PDF
  • Word
  • MP4
  • CSV
  • Shapefile
  • DXF
The online stopping sight distance report: the LiDAR scan rendered from the driver's eye with map and zoom insets, beside a table of every station's visibility, obstruction, milepost, slope and maximum safe speed
The online report: one page per road and direction, every tested point on it (17,274 on this corridor). Pick a row and the scan jumps there: the driver’s eye view, the target it was measured against, a zoom on whatever is in the way and the point on the map.
  • Summary reportThe corridor's deficiencies, ranked, with the criteria each one failed.
  • Comprehensive reportEvery station of every analysis, the numbers behind each verdict and the standard applied.
  • One per road and directionSplit the way a project is reviewed, in PDF and in Word.

Connect the findings to the crashes

A deficiency is only half the story. The Safety Dashboard puts a corridor’s findings next to your state’s crash records, on the same mileposts, so a run of blocked sight distance becomes a place where drivers are crashing.

The Safety Dashboard on a monitor in an office: four corridors on a map with stopping sight distance deficiency runs, crashes by severity, crash clusters and intersection rings, beside a panel of crash counts and expected savings
  • 88

    Crashes in the filter

    on 43.27 mi of selected corridor

  • 2.03

    Crashes per mile

    four corridors of SR-102 and SR-23

  • 1 / 13 / 73

    Fatal / injury / property damage

    severity split of those crashes

  • 12

    Stopping sight distance runs

    0.47 mi — 1.1% of the corridor

  • 10%

    Crashes beside a run

    within 0.1 mi, four of them inside one

  • $984K / yr

    Expected savings

    if the priced countermeasures are built

Four corridors of SR-102 and SR-23, stopping sight distance against the state’s 2021–2026 crash data. Twelve deficiency runs cover 1.1% of the 43.27 selected miles — and 10% of the crashes fall within a tenth of a mile of one.

RDV’s safety engineers work it with you: choose the analysis, the direction and the corridors, and the dashboard answers whether crashes really concentrate at the deficiencies, then prices what fixing them would be worth.

  • Ask one question at a time

    Filter to an analysis, a direction, a milepost range and the corridors you care about. The map, the counts and the benefit figures all follow the filter.

  • Both layers on one map

    Deficiency runs drawn by severity, crashes by severity — fatal, injury, property damage, animal strike — plus crash clusters, intersection sight-distance rings and glare-time matches.

  • Milepost by milepost

    Under the map each corridor breaks down in route order: the crashes, the deficiency runs and the intersections on every stretch.

  • A statistical verdict, not a hunch

    A permutation test against random placement on the same corridor says whether the crashes really do concentrate at the deficiencies, per analysis.

  • A priced countermeasure program

    Crash modification factors from the FHWA CMF Clearinghouse and FHWA crash costs give crashes prevented, annual benefit and a benefit-cost ratio for every site — the numbers an HSIP application asks for.

  • Intersection priorities

    Every junction in the state data ranked by comprehensive crash cost, or by what a countermeasure there would be worth.

Crash data comes from the state’s own dataset, matched to your corridors by milepost. One state is connected today; others follow as their data is loaded.

No LiDAR? No problem

If the road exists only as a design, the audit still runs. RSA 3D began on CAD models — sight distance checked on proposed geometry — and only later learned to read LiDAR of roads already built. Both still run on the same platform, against the same standards.

  • Send what your designers already work in

    Bentley MicroStation and OpenRoads, Autodesk Civil 3D and InfraWorks, SketchUp, or a plain DGN or DWG. We build the 3D model from the design files themselves — nothing has to be redrawn for the audit.

  • The same analyses, the same standards

    Stopping, passing and intersection sight distance, curves, clear zone and the rest, measured on the design surface against the same AASHTO criteria a scan is held to.

  • Findings while they are still cheap to fix

    A sight line that fails on the model costs a design revision. The same failure found after construction costs a change order, a retrofit — or a crash record.

  • Alternatives, compared before anyone commits

    Run each alternative and see which geometry actually drives safely, then take the answer to the design review with the model to show for it.

A design run hands back the same reports, video and tables, plus a colour-coded CAD drawing of the maximum safe speed at every location — in a form the design team can take straight back into the model. See it on the Henderson Interchange.

A proposed interchange design loaded into Road Safety Audit 3D
Sight distance checked on a model built from the design files

Built for the whole network

Analyses follow the AASHTO Green Book (7th edition), the AASHTO Roadside Design Guide and the MUTCD.

  • Cloud scale

    A run fans out across up to 180 parallel jobs a stage, with GPU rendering for visibility. Reruns reuse the prepared scan instead of processing it again.

  • 3D viewer in the browser

    Stream the project in full 3D with saved views, environment settings, layers, animations and virtual tours.

  • Routes on the map

    Draw paths and lanes over the scan on a map, or build routes from a road and milepost range where state milepost data is connected.

  • Secure sign-in

    Two-factor sign-in with an authenticator app, and roles that keep each manager and user to their own projects.

Run your first corridor in 123BIMLive

Send us a LiDAR scan of one road and see every finding for it in your own project.