We make use of a wide variety of data sources to answer questions about the role of animals in shaping landscapes and animal responses to landscape changes. These range from satellite and UAV-based (aka drone) remote-sensing to fine-scaled GPS telemetry and field surveys. Below are descriptions of some of the technologies we have available.
Images courtesy of Evan Hockridge
The Harvard Animal Landscape Observatory (HALO)
The HALO is a custom designed multi-sensor remote sensing platform built by Phoenix LiDAR Systems. Consisting of an industry leading LiDAR unit, a thermal imager, and a RGB camera integrated together with a centralized data acquisition computer, the HALO is exceptionally capable of collecting high-resolution remotely-sensed data over large spatial extents. Weighing 5 kg and with multiple mounting points, the HALO is designed for integration on an unmanned aerial vehicle (UAV), but can also be mounted on manned aircraft, terrestrial vehicles, tripods or personnel carried backpacks, allowing for maximum versatility in data acquisition.
The primary sensor in the HALO is a Riegl VUX-1LR Light Detection and Ranging (LiDAR) sensor integrated by Phoenix LiDAR Systems. This LiDAR unit is the current gold standard for high-resolution low-altitude airborne laser scanning and is likely the most capable UAV mountable laser scanner available. Being the long range (LR) version of the VUX-1, this sensor can collect high resolution data at altitudes up to 530 m and horizontal scanning distances up to 1 350 m. The spinning sensor design allows for an unrestricted field of view approximating 360 degrees and a rotational speed of up to 200 rotations per second. These factors, in addition to the pulse repetition rate of up to 820 KHz, enable the sensor to be flown by any number of aircraft types at various altitudes and speeds while collecting more than 750 000 measurements per second. The result of the sensor performance is three-dimensional data at resolutions that allow highly detailed analysis of fine-scaled landscapes features across large spatial extents.
The HALO utilizes a heavily modified Tau 2 thermal imaging sensor constructed by FLIR. The modifications result in an integrated and precise thermal imaging payload. An external frame has been built around the base imager, which, along with additional software enhancements, allows each exposure to be timestamped precisely, a necessity for georectification the thermal data. Vignetting, a common issue for thermal imagery, is corrected using a mechanical shutter built into the frame that replaces the rolling shutter built into the sensor body. The critical issue of thermal drift is corrected for using a NUC/FFC calibration that is performed multiple times per flight. Other additional components, such as cooling fans, have also been added to the sensor to solve common issues with thermal remote sensing. In tests post modification, the sensor was able to measure object temperatures within 1.5 degrees C of truth on structured surfaces and has an estimated performance of 3 degrees C of truth on unstructured surfaces with correct calibration. While precise temperature mapping is not usually required for animal identification or the identification of features that show clear temperature differences within landscapes, the ability to measure absolute temperature allows a nuanced understanding of how temperature drives animal decisions and enables the construction of models to delineate landscape features more precisely than non-imaging thermal cameras allow.
A modified Sony A6000 camera is utilized on the HALO to provide RGB color images. All unnecessary parts of the camera have been removed to leave only the sensor body, a fixed focal prime lens, an external trigger for testing purposes, and a custom case much smaller than the original. The sensor is ideal for high resolution RGB mapping of landscapes due to the large APS-C sensor dimensions, mechanical shutter, 24.3MP resolution, and a mirrorless sensor shutter speed. These traits set the sensor apart from most RGB imaging platforms because although it is a similar size to typical drone cameras, it has a significantly higher resolution and is not reliant on a rolling shutter standard in many commercial UAV cameras.
The HALO is integrated with Phoenix LiDAR System’s next generation AIR Navigation Box (nav box). Containing all additional systems needed for multi-sensor integrated data collection, AIR is significantly lighter than other nav boxes used on the Ranger-LR sensor, effectively making our LiDAR sensor a Ranger-LR Lite. Weighing just 450g, HALO’s nav box contains a powerful MEMS IMU, onboard computer with custom acquisition and integration software, inputs and outputs for each sensor, and inputs for a PPK/RTK GPS unit. During integration, the exact position of each sensor type, the nav box, and any attached GPS is precisely mapped within the payload and provided to the nav box computer. By having this information, unparalleled co-registration of each sensor type is possible, and thus the data collected by each sensor is easily associated with those collected by the other sensors. Information from the IMU and GPS antenna geo-rectify the imagery and LiDAR point data in real time.