You may read this in a copy of the magazine that you picked up at the Geo Week conferences in Denver on 13-15 February 2022. Be inspired by growth of the event and its vibrant exhibition, with more than 175 companies, most of which are in UAVs, cameras, lidar, photogrammetry, software and geospatial services. Exciting, vibrant, expanding!
This editorial, however, was written when a less vaunted event was in progress. The Point Cloud Processing Workshop took place in Stuttgart on 26-27 January, organized jointly by the University of Stuttgart and EuroSDR. The program featured speakers entirely from the European mainland, giving a stimulating series of updates and insights, from the academic world, government agencies and research institutes, and software suppliers. AI’s central position in modern lidar was confirmed in one session, “Semantic Segmentation—State-of-the-art and Generation of Training Data.” I’m intrigued by the title of one of the papers therein, “Casting the paid crowd and the machine into a hybrid intelligence system for 3d point cloud segmentation,” by Michael Kölle, who is a co-author of an article that will be in the next issue of this magazine. Deep learning, hard to understand though it may be, actually works and is penetrating more and more areas of life; lidar processing is on a springboard, ready to jump to much higher levels of automation in the extraction of information from point clouds. We have become accustomed to impressive improvements in lidar hardware over the years—both automotive and airborne lidar are testament to this—but deep learning will not disappoint as it becomes standard in off-the-shelf-software. The proceedings of this meeting will be essential reading.
We bring you three articles in this issue that are important in different ways. Following our trip to Germany last fall, we have tried to lay to rest the sad story of Martin Isenburg and move the spotlight to focus on the resurrection of his company, rapidlasso GmbH. We hope to have preserved Martin’s legacy while at the same time highlighting the tragedy of mental illness. Equally, Martin’s customers and prospects will want to know that the products are not only available but also evolving.
Secondly, our periodic peregrinations into radar have unearthed a company in San Diego, which uses L-band to detect water leaks. Given the scarcity of water in vast areas of the globe and the potential to avoid tragedy where water leaks are a portent of failure in dams or seawalls, this matters. The article also contains insights into managing a startup, i.e. creating a sustainable business while working on perfection of the underlying idea at the same time.
Lastly, we feature a Romanian survey company using a UAV system with both lidar and cameras from a Swiss company to map a Romanian ski resort and its environs for planning purposes. The lesson here is that we Americans, while proud of the incredible lidar integrators successfully doing business in the US, must be conscious that there’s a lot going on elsewhere in the world. The YellowScan event last year was a vivid example1.
I’ve lamented in previous editorials the lack of space to bring to you gems from the recent literature or online. Let’s squeeze in a few.
Our world today is burdened by conflict and warfare. Ukraine is one of many. Some wars, nevertheless, come to an end, but the work of recovery is daunting. Mine-clearance is one of the riskiest parts, but without it normal civilian life, especially agriculture, cannot recommence safely. It’s heartening to know that the HALO Trust, a British mine-clearance charity, has been using UAVs equipped with thermal sensors and lidar to detect landmines in Angola2. Other sophisticated approaches, for example ground-penetrating radar, are available too and it is heartening that technologies used to generate geographic information have immediate, successful humanitarian applications too.
Do you sometimes become irritated when starting to read a paper because the authors cite far too many references just in the introduction? Are they showing off their erudition, their mastery of the field, do they want to ensure they mention works by people who may be influential in their future funding, or is that too cynical a view? Perhaps. Our ability to create knowledge and store it for future generations seems boundless: “The more knowledge humans have, the longer it takes a budding researcher to get to the frontier, and thus to push things forward.”3 Ideas are becoming harder to find. Different models for funding research, therefore, could be the answer. DARPA could be a clue. The article from which the above quotation was drawn applies not specifically to geospatial research but all across the sciences. Worth pondering.
Let’s end by moving from this worrying, but rather philosophical thought about science to recent technological developments. Most of us have heard about the merger of Velodyne and Ouster4. The CEO of the new operation will be Angus Pacala—you’ve read his thoughts on these pages!5 Meanwhile, a major competitor, Luminar, has announced that the Volvo EX90 electric SUV is the latest car to have its lidar fitted as standard, claiming a range of 600 m6. These firms—and their competitors—could be helped in the future by researchers at the Fraunhofer Institute for High Frequency Physics and Radar Techniques, who have announced a new approach to combining multiple sensors into a car’s headlight7. Enabling lidar and radar to follow the same path as the LED light is no mean feat. Although we go to our traditional airborne lidar suppliers for high-end, high-accuracy, low-noise UAV-lidar systems, we benefit enormously from the opening up by the automotive lidar firms of this geospatial market space.
Thus we’re well into 2023 and lidar seems poised to make great advances. This year the remarkable USGS 3DEP program will comfortably exceed 90% in its national coverage, a triumph not only of securing funds but of managing a host of partners all committed to using rapidly changing technologies to best advantage. We’re excited about the next generation: “The 3D National Topography Model (3DNTM) is a new initiative that updates and integrates USGS elevation and hydrography data, and the relationships between them, into a 3D model to deliver higher-quality data and support improved geospatial analysis.”8 At Geo Week, at events further into the calendar, online and in the daily life of suppliers, services companies and end users, lidar will become more pervasive. Thank you for reading LIDAR Magazine. It’s an honor to bring you news about how our technology matters.
2 Anon, 2022a. Minecraft: modern technology is helping to reduce the menace of landmines, The Economist, 445(9324): 44, 3 December 2022.
3 Anon, 2022b. In search of a bright light: billion-dollar experiments aim to end a period of scientific stagnation, The Economist, 445(9319): 72, 29 October 2022.
5 Walker, A.S., 2022. Digital lidar for everything, LIDAR Magazine, 12(3): 4-16, Fall 2022.
7 Anon, 2023. Space-saving sensors ease automotive design woes, Photonics Spectra, 57(1): 32-33, January 2023.
8 usgs.gov/national-hydrography/3d-national-topography-model-call-action-part-1-3d-hydrography-program#:~:text=The%203D%20National%20Topography%20Model%20(3DNTM)%20is%20a%20new%20initiative,and%20support%20improved%20geospatial%20analysis.