LiDAR sensors are consistently deployed across various industries to capture accurate spatial representations of variable target feature sets. It could be as simple as performing a wide-area acquisition of a flood plain to generate an elevation model (DEM/DTM) or a more complex, multi-stage survey of a utility distribution network for vegetation encroachment analysis. Regardless of the survey specifics, GNO-SYS has found that clients succeed best when they follow these steps. It’s also a great idea to spend some time up front thinking about how you’ll organize your data.
Acquire & Extract
At this stage, the raw LiDAR scan data is acquired by one or more acquisition platforms at a variable cadence (daily, per-flight, etc.) and the resultant data payloads are extracted from the platforms. At the end of a given acquisition day, there can be one or many extracted data payloads to be managed.
Acquisition crews are generally busy with tracking survey time, cost, and schedule so implementing an automated data extraction process to produce consistent data payloads greatly reduces risk of error or data loss.
As soon as is possible – the newly acquired LiDAR data must be processed from its ‘raw’ state into a ‘ready’ state (ASPRS LAS/LAZ) so that the quality of the data can be assessed. Once the acquisition crews have ensured the newly acquired data meets specification, the resultant data payload can be transmitted offsite for processing.
This step is critical to the success of a survey – if acquisition crews can detect acquisition failures within a short time frame, they are able to issue re-flights prior to mobilizing away from the area-of-interest (AOI) which greatly reduces costs and time slippage.
Once the ‘ready’ data has been validated to meet quality specification and transmitted offsite by the acquisition crew it is now possible to perform automated processing steps to prepare and ingest the LiDAR data into the client’s production pipeline.
Generally, these steps are domain-specific such as ground extraction, base feature classification, or splitting of LAS/LAZ data into a grid index. Regardless, they generally are consistent within a single survey and should be automated to ensure maximum consistency of LiDAR data coming into the production pipeline.
Now, the LiDAR data has already been prepared into the best-possible ‘ready’ state and ingested into the client’s domain-specific production process. Whether the production is entirely automated, manual, or a combination of the two it is imperative that incoming LiDAR data is in a consistent and predictable state. Any deviation from the expected input can (and will) break downstream processes and introduce significant risk to the success of the survey.
Similarly, many processes cannot be completed piece-meal and require an entire region of data be available prior to start. Ensuring data becomes available in a consistent manner allows for automated checks for coverage and teams can be alerted as soon as sufficient data is made available.
Once the data is available and production teams are alerted of availability, the client is now able to derive their domain-specific intelligence and value-added data products while ensuring maximum visibility into the status of the production process without significant management overhead.
As the production process progresses, the client can act upon the derived intelligence and data products as quickly and reliably as possible which provides value not only to them but to any of their consumers.
Using vegetation encroachment as an example, some consumers require notification of any encroachments within a specific distance (called minimum-vegetation-clearance-distance or MVCD) within 24 to 72 hours of acquisition completion. Any disruption to upstream stages will have significant impact on this requirement and can make or break a survey.
A LiDAR survey doesn’t have to be complex or difficult, and using these steps can help streamline the workflow whether you’re doing aerial or terrestrial lidar.