A stopped vehicle inside a tunnel, a worker entering an equipment path, debris in a rail corridor, bridge movement after a flood, and slope movement during heavy rainfall all share the same monitoring problem: the information is most valuable when it can be measured, interpreted, and acted on quickly.
Real-time LiDAR monitoring uses three-dimensional measurements to track infrastructure conditions and activity around infrastructure. Unlike a traditional LiDAR survey that documents conditions at scheduled intervals, a real-time or near-real-time system reduces the delay between observation and response.
The value of real-time LiDAR monitoring is not the continuous collection of point clouds, but the ability to turn spatial measurements into timely information such as an alert, inspection ticket, clearance status, maintenance priority, or engineering review.
Real-world use varies by application area. Roadside LiDAR has already been demonstrated for traffic, intersection-safety, and work-zone monitoring [1]–[3]. Other applications, including construction comparison, clearance monitoring, bridge review, dam monitoring, and slope stability, range from periodic and event-triggered measurement to specialized real-time monitoring systems [4]–[7].
Not every application discussed in this article is currently deployed as a continuous real-time system; where documented deployments are limited, the application is identified as near-real-time, event-triggered, emerging, or potential.
Table of Contents
Why LiDAR Is Useful for Monitoring
LiDAR is useful in infrastructure monitoring because many infrastructure questions are geometric. Operators and engineers often need to know whether something moved, whether an object is inside a clearance envelope, whether a constructed element matches the model, whether a slope surface changed, or whether a person or vehicle entered a restricted zone.
A LiDAR point cloud can support measurements of distance, height, alignment, displacement, clearance, movement, and spatial relationships.
LiDAR is one of several technologies used for infrastructure monitoring. Cameras provide visual information, radar measures motion and speed, and sensors such as GNSS, strain gauges, and accelerometers can measure conditions at specific locations. LiDAR is particularly useful when the shape, position, movement, or spatial relationship of objects needs to be measured across an area.
In practice, real-time infrastructure monitoring often works best as a sensor-fusion problem. LiDAR can provide geometry, cameras can provide visual appearance, radar can contribute speed or motion information, GNSS and inertial measurement units can support localization, and asset databases can provide reference information.
What Real-Time Monitoring Means
Real time does not have one universal latency requirement. The correct update rate depends on the decision being supported. A collision-warning system may need a response in seconds or less. A traffic operations dashboard may need updates every few seconds or minutes. A construction progress system may only need updates every few hours. A slope or bridge monitoring system may not need continuous scanning during normal conditions, but may need higher-frequency monitoring after rainfall, flooding, excavation, impact, or abnormal movement.
| Monitoring mode | Typical cadence | Example use cases | Best description |
| Immediate real-time monitoring | Milliseconds to seconds | Work-zone intrusion, tunnel obstruction, pedestrian-vehicle conflict, rail crossing hazard | Real-time safety monitoring |
| Operational near-real-time monitoring | Seconds to minutes | Intersection analytics, queue monitoring, traffic incident detection, facility movement | Near-real-time operations |
| Decision-support monitoring | Minutes to hours | Construction progress checks, scan-vs-BIM comparison, post-event bridge scan, slope movement review | Near-real-time decision support |
| Event-triggered monitoring | Activated after a defined event | Flood, storm, vehicle impact, excavation, abnormal reading, emergency inspection | Event-triggered monitoring |
| Routine monitoring | Hours to months | Asset inventory, corridor mapping, scheduled tunnel scan, annual bridge or dam documentation | Routine monitoring, not usually real time |
Three Monitoring Classes
Real-time LiDAR applications can be grouped into three main classes: operational perception, geometric compliance and clearance, and deformation or structural monitoring.
| Monitoring class | Main question | Common examples | Typical outputs |
| Operational perception | What is moving, where is it moving, and does it create a safety or operating issue? | Vehicles, pedestrians, cyclists, workers, equipment, queues, facility movement | Object tracks, conflict alert, intrusion alert, queue length, incident flag |
| Geometric compliance and clearance | Does the built or operating condition satisfy a geometric rule? | Tunnel clearance, rail clearance, loading-gauge compliance, scan-vs-BIM comparison, construction tolerances | Clearance status, deviation report, missing component, quality-control exception |
| Deformation and structural monitoring | Has an asset or surface moved, shifted, deformed, or changed over time? | Bridges, dams, levees, slopes, retaining walls, tunnels, rock faces | Displacement estimate, deformation trend, surface-change map, engineering review flag |
How a Sensing-to-Decision System Works
A real-time LiDAR monitoring system can be described as a sensing-to-decision pipeline. Although systems vary by application, they generally collect LiDAR data, process it, identify relevant changes or objects, and produce information for operators or engineers.
AI can be used to identify and track vehicles, pedestrians, workers, or equipment. It can also help compare scans with digital models or identify unusual changes in infrastructure.
AI should be described as an enabling layer, not as a replacement for engineering judgment. In safety-critical applications, AI results still need appropriate validation and, where necessary, human review.
| Step | Purpose | Example output |
| Capture | Collect LiDAR measurements | Point cloud |
| Align | Place measurements in a common reference frame | Aligned data |
| Analyze | Detect objects, movement, or changes | Tracks or measurements |
| Compare | Check against rules or previous conditions | Deviation or alert |
| Validate | Check reliability | Confirmed or rejected alert |
| Act | Send useful information to users or systems | Alert, report, or work order |
Edge computing is also important because point-cloud data can be large. Sending every raw frame to a distant server may be expensive, slow, or unreliable. Edge processing allows the system to process data near the sensor and transmit only useful outputs, such as object tracks, event clips, deformation metrics, compressed point clouds, or alerts. The tradeoff is that edge systems can reduce communication and storage requirements, but they add hardware and maintenance needs.
Where It Is Used
Operational perception is the clearest current real-time use case. Roadside LiDAR has been studied and demonstrated for traffic participant detection, tracking, and cooperative driving automation [11]. The Cyber Mobility Mirror project in Riverside, California, used roadside LiDAR to detect and track road users at a real intersection [1]. PRISA was tested at a signalized intersection in Chattanooga, Tennessee, where LiDAR was used to assess traffic-safety conditions [2].
Work-zone safety is another area with real research activity. Safe-D researchers have also tested roadside LiDAR for monitoring vehicles and workers in active work zones [3]. This type of system could support warnings for workers, supervisors, or equipment operators in active work zones.
Geometric compliance and clearance applications are often near-real-time, periodic, or event-triggered rather than continuous. Rail corridors, tunnels, construction sites, and work zones all depend on spatial rules. In construction, as-built point clouds can be compared with BIM or IFC models to identify missing, incomplete, or misaligned components [4]. A documented railway example comes from the Shuanghekou Tunnel on the Chengdu–Kunming Railway in China, where mobile laser scanning was used to reconstruct tunnel geometry and compare it with railway clearance requirements [13]. The study demonstrated that LiDAR-derived point clouds could identify clearance conditions with a reported inspection precision of approximately 0.03 m. This example illustrates how LiDAR can support geometric compliance without requiring continuous real-time operation. Similar rail and tunnel applications can support clearance checks, obstruction detection, and profile comparison, but they should be described as periodic, near-real-time, event-triggered, or emerging unless continuous operation is specifically documented.
Deformation and structural monitoring includes bridges, dams, levees, slopes, retaining walls, rock faces, and tunnels. LiDAR is widely used for periodic or event-triggered geometric measurement in these areas, while continuous real-time use is more specialized [5]. Slope stability is one area where real or near-real-time LiDAR-related systems are more visible. Maptek Sentry is a commercial laser-scanner-based monitoring system used for mining and civil engineering projects, including road cuttings, landslips, dam walls, and tunnel openings [6]. Research examples also exist, including high-resolution terrestrial laser scanning tested on the Hochebenkar rock glacier landslide in the Eastern Austrian Alps [7].
For bridges, dams, tunnels, and slopes, LiDAR should not be presented as a standalone structural diagnosis tool. It measures geometry and surface change, but it does not directly measure internal stress, hidden corrosion, material strength, foundation condition, or load capacity. Engineers still need to determine whether the measured change is structurally meaningful. In these settings, LiDAR is best understood as one layer in a broader monitoring program that may also include visual inspection, hydrologic data, geotechnical instruments, structural sensors, and professional review.
Some applications remain future-looking or hypothetical. A facility-wide LiDAR system could monitor service-route blockages, equipment movement, restricted-zone entry, or occupancy patterns. A dam or slope system could increase scan frequency during storms or abnormal readings. Unless a specific deployment is documented, these should be framed as potential or emerging applications rather than established practice.
When Real-Time LiDAR Is Not Needed
Real-time LiDAR is not appropriate in all monitoring scenarios. It may not be justified when an asset changes slowly, delayed detection has low consequence, periodic inspection already provides enough information, or a simpler sensor can answer the same question at lower cost.
It may also be unnecessary when a site lacks reliable power, networking, mounting locations, calibration access, or maintenance support. Even if the technology works, it may not create value if alerts are not owned, reviewed, prioritized, or connected to response procedures.
How Performance Should Be Evaluated
Performance should be evaluated across both technical and operational dimensions, with the appropriate measures depending on the application. Common measures include accuracy, false alarms, response time, geometric measurement accuracy, system reliability, and performance under different environmental conditions.
Operational measures are just as important. Did the system produce actionable alerts? Did it reduce the inspection burden? Did it improve response time? Did it support better maintenance prioritization? Did it provide useful evidence after an incident? Did it reduce exposure for workers or inspectors?
AI performance requires careful evaluation. Models may perform differently when deployed at a new location because sensor placement, traffic conditions, weather, infrastructure, and other site characteristics can change. Systems therefore need to be tested under the conditions in which they will operate.
Deployment Risks and Limitations
Real-time LiDAR systems have technical, financial, and organizational challenges. Deployment can require substantial supporting infrastructure, including power, communications, computing, data storage, calibration, and maintenance.
LiDAR also has physical limits. LiDAR measurements can be affected by occlusion, weather, range, surface properties, sensor placement, and vibration. Calibration drift and registration error are especially important for change detection, deformation monitoring, and scan-vs-BIM comparison.
Data management is another challenge. Mobile and stationary LiDAR datasets can be large, especially in multi-sensor or high-frequency deployments. NCHRP guidance on mobile LiDAR notes that the size and complexity of laser scan data can create storage and information-technology challenges if systems are not prepared for the volume [12].
Privacy should be treated as a design issue. LiDAR can offer privacy advantages compared with camera-only monitoring because it primarily captures geometry rather than detailed facial appearance or color imagery. However, LiDAR is not automatically privacy-free. High-resolution spatial data can reveal presence, movement patterns, repeated behavior, and trajectories, especially when combined with cameras, device identifiers, or long-term tracking. Responsible deployment should include purpose limitation, retention rules, access control, cybersecurity, public transparency, and aggregation of outputs where possible.
Conclusion
Real-time infrastructure monitoring with LiDAR represents a shift from passive documentation toward active infrastructure awareness.
The most useful applications fall into three groups: operational perception, geometric compliance and clearance, and deformation or structural monitoring. These groups differ in latency needs, baseline requirements, technical maturity, and deployment risk. Traffic and work-zone perception are already being demonstrated in real-world environments, while construction comparison, clearance monitoring, and deformation monitoring are often near-real-time, event-triggered, specialized, or integrated with broader inspection workflows.
Successful LiDAR monitoring depends on clear thresholds, reliable analytics, appropriate human review, and established response procedures. When these elements are properly integrated, LiDAR can support safer infrastructure operations, more targeted maintenance, and faster response to changing conditions.
References
[1] Bai, Z., Nayak, S. P., Zhao, X., Wu, G., Barth, M. J., Qi, X., Liu, Y., Sisbot, E. A., and Oguchi, K. “Cyber Mobility Mirror: A Deep Learning-Based Real-World Object Perception Platform Using Roadside LiDAR.” IEEE Transactions on Intelligent Transportation Systems, vol. 24, no. 9, 2023, pp. 9476–9489. DOI: https://doi.org/10.1109/TITS.2023.3268281
[2] Bang, T., Abubakr, H., de la Garza Villarreal, E., Nguyen, T. P., Harris, A., Hirano, T., Sartipi, M., Xu, Y., and Nguyen, H. H. “PRISA: Proactive Infrastructure LiDAR Framework for Intersection Safety Assessment.” arXiv, 2026. URL: https://arxiv.org/abs/2607.16156
[3] Safe-D National UTC. “Development of a Roadside LiDAR-Based Situational Awareness System for Work Zone Safety: Proof-of-Concept Study.” Final Research Report, 2023. URL: https://rosap.ntl.bts.gov/view/dot/73278
Supporting dataset DOI: https://doi.org/10.15787/VTT1/KUTAZD
[4] Kavaliauskas, P., et al. “Automation of Construction Progress Monitoring by Integrating 3D Point Cloud Data with an IFC-Based BIM Model.” Buildings, vol. 12, no. 10, 2022, article 1754. DOI: https://doi.org/10.3390/buildings12101754
[5] Kaartinen, E., Dunphy, K., and Sadhu, A. “LiDAR-Based Structural Health Monitoring: Applications in Civil Infrastructure Systems.” Sensors, vol. 22, no. 12, 2022, article 4610. DOI: https://doi.org/10.3390/s22124610
[6] Maptek. “Sentry: Laser-Based Stability and Surface-Movement Monitoring.” URL: https://www.maptek.com/products/sentry/index.html
[7] Hosseini, K., Zubareva, S., Hummelsberger, J., and Holst, C. “Improved 4D Feature-Based Deformation Tracking for High-Resolution Real-Time Landslide and Slope Deformation Monitoring Based on Terrestrial Laser Scanning.” Natural Hazards, vol. 122, no. 5, 2026, article 178. DOI: https://doi.org/10.1007/s11069-025-07939-0
[8] Federal Highway Administration. “Digital As-Builts.” U.S. Department of Transportation. URL: https://www.fhwa.dot.gov/construction/dabs/
[9] Federal Highway Administration. “Advanced Digital Construction Management Systems.” U.S. Department of Transportation. URL: https://www.fhwa.dot.gov/construction/adcms/
[10] Federal Highway Administration. “Integrated Digital Project Delivery.” Every Day Counts, U.S. Department of Transportation. URL: https://www.fhwa.dot.gov/innovation/everydaycounts/edc_8/digital_project_delivery.cfm
[11] Sun, P., et al. “Object Detection Based on Roadside LiDAR for Cooperative Driving Automation: A Review.” Sensors, vol. 22, no. 23, 2022, article 9316. DOI: https://doi.org/10.3390/s22239316
[12] National Cooperative Highway Research Program. Guidelines for the Use of Mobile LIDAR in Transportation Applications. NCHRP Report 748, Transportation Research Board, 2013. URL: https://www.trb.org/Publications/Blurbs/169111.aspx
[13] Zhou, Y., Wang, S., Mei, X., Yin, W., Lin, C., Hu, Q., and Mao, Q. “Railway Tunnel Clearance Inspection Method Based on 3D Point Cloud from Mobile Laser Scanning.” Sensors, vol. 17, no. 9, 2017, article 2055. DOI: https://doi.org/10.3390/s17092055
