Accuracy – Efficiency – Consistency. Requirements for Today’s Best Lidar Sensors

This interview provides an inside perspective of NV5, a leading provider of technology, conformity assessment and consulting solutions for public and private clients worldwide. It is being provided here as told by the persons being interviewed. 

Accuracy – Efficiency – Consistency. Requirements for Today’s Best Lidar Sensors

Background

Commercial lidar entered the geospatial services arena in the 1990s, becoming a viable technology with the maturation of Global Positioning (GPS) and the introduction of inertial navigation technologies. NV5 entered the market almost immediately. Our first sensors had lasers capable of 2 KHz pulse rates, which pales in comparison to the technology we use today. But it was an impressive start.

We never looked back, quickly growing to one of the world’s largest and most capable lidar service providers today. We have constantly upgraded our airborne sensors over the last 25 years, added ground-based mobile mapping capabilities in the mid-2000s, and adopted topo-bathymetric lidar when it was first introduced by RIEGL in 2013. We have diversified our acquisition platforms, adding rotary wing and unmanned systems to our fixed-wing aircraft. We began integrating robotic systems that combine lidar with visible and thermal IR cameras to carry out remote inspections of isolated sites just a few years ago. Lidar plays a key role in the services we offer our clients.

Technology Overview

Our sensors are impressive. So too is the collection of software that we use to turn the sensor data into actionable intelligence. For terrestrial data acquisition, we rely solely on sensors manufactured by RIEGL. Fixed-wing acquisition is primarily captured with one of four VQ-1560ii-S or two VQ-780ii-S sensors. These date to 2017 for the 1560ii and just a year ago for the 780ii models. Rotary-wing and unmanned platforms typically carry one of the RIEGL VUX lidar sensors. We were one of the early adopters of VUX technology when it was first introduced in 2013. We also use the RIEGL VMX-2HA mobile mapping technology.

For bathymetric operations, one of the sensors we use is the RIEGL VQ-880gii. Again, we were very early adopters of this specialized technology, purchasing multiple sensors almost immediately when the first commercial sensors were introduced.

Our suite of software includes many contributors such as RIEGL post-processing, popular off-the-shelf modules from TerraSolid, BayesMap Solutions, Esri, and many custom applications developed internally by our R&D engineers.

Technology Selection Process 

We have worked closely with all lidar manufacturers over the years, and owned sensor technology at different times from all providers. Recently, however, we have focused almost exclusively on RIEGL sensors because they have the best performance characteristics for the wide range of projects that we undertake. 

When assessing technology, we evaluate many factors. This includes the ability to achieve very high point densities efficiently, with precise three-dimensional accuracies. We also need the ability to penetrate dense vegetation canopies, reach greater depths in bathymetric applications, adapt to both urban and rural project environments, and perform equally well in flat and steep terrain. The reliability of the sensor is critical as the real costs of losing a sensor for two to three weeks during acquisition can be disastrous to our best-laid project plans. Finally, the downstream workflow for moving the raw data to intelligent point clouds must fit within our production environment.

Systems Integration

RIEGL’s consistency in utilizing Ri-Acquire & Ri-Process across their full suite of sensors enables us to seamlessly integrate any RIEGL sensor into our established workflows. There are nuances to learn with the deployment of any new sensor, including identifying and tailoring sensor settings specific to each application. Additionally, there will always be lessons learned when deploying new technology and sometimes those lessons come the hard way! We leverage the insights gained to continually evolve and advance our workflows and best practices. Advancements can be as simple as a small tweak to the applied sensor settings, a refinement in our processing algorithms, or as big as rethinking our entire approach to data acquisition based on the latest sensor developments. Ultimately, the breadth of RIEGL sensors that we own and operate allows us to offer our clients the best solutions available. 

Project Snapshots

Yosemite

Lidar is a powerful remote-sensing technology that is extremely useful for developing a clear understanding of forests by analyzing detailed elevation models that stretch from the treetops to the ground. These models have considerable detail beyond tree heights. They also contain valuable information on canopy densities and the distribution of branches throughout. The innovative scan pattern of the RIEGL sensors with a slight forward and aft look of the two scanning mirrors is ideal for success in heavy vegetation. The complete data allows forest managers and other scientists to develop a keen understanding of past fires, diagnose vulnerabilities, and improve forest resilience.

In 2019, NV5 Geospatial worked with researchers at the Forest Resilience Laboratory and the University of Washington, Seattle, and the US Geological Survey to capture a high-resolution elevation model of Yosemite National Park in the Sierra Nevada Mountains of California. These data were captured with a RIEGL 1560ii sensor at maximum densities of 60 points per square meter (ppsm).

Khumbu Glacier

The Khumbu Glacier lies at the base of Mount Everest in Nepal. It has the distinction as the highest glacier on earth, reaching 25,000 feet on the Lhotse Face. Everest Base Camp lies along the side of the glacier and is the starting point for summit attempts. In 2019 the National Geographic Society looked to NV5 Geospatial to create a point-in-time detailed 3D model of the glacier to fuel scientific analysis on climate change and the impacts on the water supply for millions of people.

High elevations provide significant challenges for helicopter operations as it is difficult for the engine to breathe thin air at increased altitudes. We collected detailed elevation models, flying only 700 feet above ground at altitudes (above sea level) ranging from 16,000 to 23,000 feet. Many sections of the glacier are almost vertical; we were able to mount the RIEGL 480ii sensor at a 45-degree look angle to ensure near-perfect data capture, especially in the rugged icefall region. We believe the capture is a new record altitude for lidar acquisition.

Archeologic Discovery in Colombia

NV5 Geospatial captured another project for the National Geographic Channel, but the challenges were significantly different from those found in Nepal. The dense jungles of the area made it impossible to capture detailed mapping of the project area with conventional remote sensing techniques. Again, we looked to a low-altitude helicopter flight equipped with dual RIEGL VUX1-LR sensors with crossing scan patterns to maximize our ability to penetrate the dense canopy and accurately map the ground below. For this project, the flight altitude was dropped to 500 feet above the ground.

These data were used by archaeologists who discovered several previously unknown settlements in Northern Colombia. They were also used by National Geographic TV in the series “Lost Cities with Albert Lin” series.

The RIEGL sensors were an important part of each of the project’s successes. For the Yosemite project, the 1560ii was ideal for rugged terrain relief, penetrating challenging vegetation, and provided cost efficiencies conservatively estimated at 20% over comparable sensors. The 480ii sensor was ideal for the Khumbu Glacier as it provided a small form factor and worked extremely well at the temperature extremes found at acquisition altitudes up to 23,000 feet above sea level. Using the small format VUX1 sensors with a crossing pattern and flying at a very low altitude allowed us to effectively penetrate dense jungle canopies. More importantly, our diverse suite of lidar sensors allowed us to select the right technology to meet each of the project’s specific challenges.

Future Use of 3D Technology

One advantage of our size and diversity of services is the ability to fund research in new technologies and solutions. Our professionals always look for a better way to meet challenges and keep a close eye on the arc of 3D technology.

Some of the advancements will be a result of improved sensor technology, adding efficiency and new capabilities. Others will come from the way we extract answers from the data. Higher resolutions and better accuracies allow us to solve problems that weren’t possible with remote sensing technology a few years ago. The way we have advanced artificial intelligence and machine learning over the last five years is even more impressive and is unlocking new possibilities. These include an increase in the frequency of remote inspection and analysis, removing boots on the ground from dangerous project environments, or introducing new solutions that improve the process for our clients and the lives of their consumers. The future is very bright.