You are here:
Field Demonstration of Unoccupied Aerial System (UAS) Waste Debris Volume Estimation
Citation:
Wussow, M., D. Durden, R. James, K. Hofacre, AND T. Boe. Field Demonstration of Unoccupied Aerial System (UAS) Waste Debris Volume Estimation. 2023 EPA International Decontamination Research and Development Conference, Charleston, SC, December 05 - 07, 2023.
Impact/Purpose:
As UAS are being adopted in a variety of ways for emergency response, we wanted to evaluate if currently available UAS sensors can render accurate volume estimations compared to currently used methods. This study evaluated four methods for obtaining volumetric data: Two UAS methods (Colorized LiDAR and Structure from Motion (RGB-SfM)) and two Ground methods (measuring tape and GNSS Points). The end goal was to compare the accuracy of volumes of debris-like materials and piles that mimic conditions found after an event requiring a clean-up response generated from UAS and ground methods, as well as assess whether UAS can provide more accurate data while reducing risk to response teams in a more efficient manner to gain a better understanding for response needs.
Description:
Battelle and EPA evaluated the effectiveness of UAV platforms to perform volumetric measurements of various types of debris (e.g., biomass such as fallen vegetation, stacks of building material, and other debris) using lidar and structure from motion (SfM) photogrammetry. The challenge of this experiment is to compare the chosen platform’s measurements to a more accurate method, which may not exist. The evaluation approach included three phases. Phase 1 included data collection of simple structures and materials of known volume, such as piles of wood pallets that can be accurately measured by hand and configured into different shapes. The pallets were used to establish a “reference cube” that could be surveyed next to varied materials in each subsequent step. Phase 2 included larger piles such as fallen tree debris pushed into a more compressed pile (with excavators) as well as raw fallen trees, stacks of recycled lumber, pipes, and other miscellaneous items. Overall, the goal was to first establish that the UAS remote sensing methodologies could accurately measure volumes of simple structures which can be validated by hand. In Phase Two, we sought to evaluate the differences in volumetric measurements of more complex piles while also establishing if UAS-based surveys are advantageous as the literature notes it should. In addition to validating remote sensing methods with hand measurements, we evaluated UAS-lidar versus UAS-structure from motion (SfM) surveys and Global Navigation Satellite System (GNSS) mapping using a rover station. The various methodologies will be evaluated against one another for accurately estimating volume, applicability in the field, and collection time Phase 3 included survey of previously un-surveyed area used as a test bed to determine the effectiveness of the platform to survey piles that had previously been unplanned and unmeasured.