A 2.228Mb PDF of this article as it appeared in the magazine complete with images is available by clicking HERE
You may not recognize the name as you have with many of the previous industry pioneers, but Jerry Dimsdale was a key player in the development of the terrestrial 3D laser scanning industry. It all began in 1992 when a fellow by the name of Ben Kacyra invited him to breakfast, but we need to lay a little groundwork before we get to that critical meeting.
The Early Years
Jerry was pursuing his Ph.D. in mathematics at UCLA in the 1970’s when he realized that most of his fellow classmates who had graduated could not find a job. Jerry recalls, "This was probably a low point in the country for employing mathematicians. Rather than spend years developing an apparently unmarketable skill, I left on an extended mountaineering vacation."
Even though he had some misgivings Jerry decided to become an engineer. He received a Ph.D. in 1982 from Berkeley in Civil Engineering, though he has never actually practiced civil engineering. Jerry stayed on at Berkeley and built several systems designed to assist in research, including the first pseudo-dynamic system for testing actual building components while predicting their performance in a large building during an earthquake by using analytical or finite element modeling for the rest of the building, and a major retrofit to the 20’x20′ earthquake simulator, increasing its fidelity, range of motion, and bandwidth.
These projects required on the job training in high-pressure hydraulics, feedback controls, multiprocessor architectures, and electrical engineering. Jerry notes, "While I didn’t realize it at the time, my appreciation of the importance of laser scanning to buildings and particularly piping systems was acquired during early efforts to design large and complicated hydraulic supply systems. Interestingly, I found that the research center at which I worked now owns two laser scanners and uses them frequently in their studies."
When asked about any events from his youth that might have influenced him Jerry recalls that while in elementary school he built a crystal radio from directions in the World Book Encyclopedia. "I remember being particularly impressed with the magic encased in a toilet paper roll wrapped with some wire, not to mention the thumb tack, razor blade, and other assorted junk, and I wondered if I could make something even better with a much bigger pile of stuff," Jerry recalls.
He continued, "I didn’t think about the crystal radio for many years until I had built a laser scanner prototype. If I could share a picture of the monstrous pile of random stuff that I had assembled it would astound you. It was then that I realized that I had just not waited long enough for the transformation to occur."
Undoubtedly the most important influence on young Jerry was his father Bernard, who was one of the pioneers of the computer industry. Bernard worked with Von Neumann as a manager on the ENIAC project. Many people consider this the first modern digital computer because it was programmable. This was actually controversial at the time, because the people building the hardware were concerned that software would slow it down–amazing.
My father was with Von Neumann when they made the now historic pitch to a general who decided whether software would be pursued. Von Neumann tasked my father with crafting the first instruction set. "Unfortunately, shortly after this my father, perhaps imprudently, defended one of his colleagues in a HUAC investigation, and was promptly considered a risk of being a communist sympathizer. To this day, ENIAC historians ignore my father’s contribution to this effort," Jerry notes.
Eventually, Bernard went on to become manager of IBM’s Los Angeles research center. During his career, he was asked to respond to a somewhat famous survey–IBM was concerned that they had saturated the market for computers with the few mainframes they had produced. My father’s unusual response was that, in his opinion, they had just barely scratched the surface.
The Cyra Days
In the early 80’s, Ben Kacyra had an AEC company called Cygna. They had a Prime computer running some CAD software, ominously called Medusa, and some analysis code that had been built in-house. Ben had been trying, unsuccessfully, for quite some time to find a way get the two to communicate. He asked a professor at UC Berkeley, Steve Mahin, if he knew anybody who could get it done. Jerry remembers, "I was randomly outside Prof. Mahin’s office at that exact moment. I had never heard of Prime or Medusa, but I was unreasonably optimistic."
Jerry spent a couple of weeks at Cygna on loan and frankly he says he got exceptionally lucky because the documentation, when available, was quite poor. At that time CAD operators were very rare and somewhat revered. Ben’s was not happy about having to show me how to start up the system. "Two weeks later he almost refused to show me again, but he became far more interested when a few minutes later I brought up a 3D dimensioned drawing of a large factory support and piping structure. I think I was as surprised as he was that it worked as well as it did, but I didn’t let him know that," Jerry noted.
Fast forward to 1992 when Ben invited Jerry to breakfast and asked him if he knew how to improve the efficiency of making as-built drawings. Jerry told him that he didn’t, but he would think about it.
Jerry recalls, "We started with a set of specifications that we thought were needed in order to meet the needs of 90% of the applications we wanted to address. Ben showed me a book that had some microwave radars in it, but after some thought that didn’t seem like the easiest idea to pursue."
Light-based sensors had shown much better resolution, but there was a big gap between measuring small objects using techniques like interferometry, and very large objects like the earth. Theoretically, there was no fundamental reason why measuring quickly at a range of 100m with high accuracy couldn’t be done, but there had been much less research in this area.
"We didn’t want to build one if we didn’t have to, particularly since at the time I knew very little about it, but we set resolution and accuracy goals that everyone in related fields told us were somewhat beyond the state-of-the-art. We scoured the country for systems, components, and software that might be suitable. Eventually, we established a company, Cyra Technologies, and I assembled a team to construct what some people regard as the first laser scanning system, "Jerry explained.
Eventually they settled on a laser that had just been invented at MIT. Jerry remembers, "After spending five months there, I came back to California with the only three working laser chips in existence. We were in negotiations with a timing vendor, but they eventually decided an impending IPO made their risk too great. I scrambled and found a bleeding-edge timing chip that had been designed at Los Alamos, but moth-balled without having ever successfully been incorporated into a system."
Their first timing board was a disappointment–while it was as accurate as any system that could be bought on the market, it was still about 8 times worse than required. Their third version of the board was finally adequate, and they had overcome a major technical hurdle.
Market penetration for the new product would take much longer than Ben and Jerry anticipated, but they were confident that the technology could have a big impact on numerous industries. The big breakthrough came when Ben approached Chevron who was concerned about how they might take on the management of several large oil fields that were largely undocumented. Jerry notes, "Despite my reluctance, Ben persuaded me to mount our ponderous contraption in my VW Vanagon. We scanned a number of Chevron facilities, and they were very impressed with the quality and quantity of the data we collected–at the time about 30 points per second!"
Cyra proceeded to manufacture and sell several different versions of the system, and eventually gained significant market acceptance. While they continued to search for the "killer application" , they became successful by working to make the technology attractive for a very wide range of applications. Ultimately, Cyra was sold to Leica Geosystems which formed the basis of their HDS laser scanning business today.
Jerry left Leica after a couple of years and at least one iteration of the same product. After a couple of years, he formed a new company, Voxis Inc, to explore a few new technical ideas. He created a new scanner with much larger range and much lower noise. Jerry also obtained patents for significant advances in flash lidar that could result in much smaller, cheaper, more accurate, and faster systems than even exist today. The scanner technology was incorporated into Topcon’s first laser scanner.
Looking Ahead
When asked for his assessment of the current state of the laser scanning industry Jerry borrows a line from his father, "While the assessment depends just a little on what you choose to include in the "laser scanning/LIDAR industry", I think it is clear that we have just barely scratched the surface."
Jerry likes to compare the laser scanning industry with the digital camera. He points out, "The digital camera which is only about 40 years old has not only become ubiquitous, but continues to expand applications at an accelerating rate. Ben told me he thought of both cameras and LIDAR’s as eyes with which the world sees itself. Myriad applications are enabled as those eyes become more acute and cost-effective."
Jerry sees at one extreme the applications emerging to support very large projects. These are happening because they represent large capital investments. Even though many project managers continue to use 2D tools because they are most familiar, he believes the economic advantages of better information will drive adoption.
He sees at the other extreme machine vision and personal applications. Phone and tablet vendors will ensure that the 3D experience is a familiar environment at very low cost, and that will drive improvements in sensors and applications. New generations of engineers will demand the benefits that can be realized. Jerry notes, "There has also been an explosion in 3D machine vision for factory automation, material processing, etc. However, the lines may soon blur between information captured by laser scanning/LIDAR, other electromagnetic frequencies, and still many other technologies such as acoustics."
Jerry believes innovation in this area is accelerating dramatically in hardware, software, and applications. However, there is currently no established standard to make reasonable comparisons between emerging technologies, and particularly no good way to assess suitability for a specific purpose. One of the most pressing needs is to be able to establish performance so that results can be relied upon. Jerry cautions, "In this regard, we are currently living in the Wild West, it is not healthy, and I don’t believe it can continue."
Jerry notes that a standards body (ASTM E57) has been in place now for over a decade, and despite lofty ambitions, has failed to introduce a meaningful way for vendors to communicate the performance of their instruments. Partly this is because some vendors want to avoid a standard that applied to their instruments could potentially impair imminent sales. Probably a bigger issue is that in most cases the usefulness of captured data is hugely dependent on the careful use of multiple instruments by experienced operators, and the eventual interpretation of the collected data for a particular use. Not only will new users find it very difficult to acquire the experience necessary to establish an adequate workflow, but also there is no effective way to evaluate resources that have already acquired that knowledge. Jerry notes, "In my experience, most everyone who sells services and/or equipment describe their products as being superlative for almost anything."
Jerry believes there is an effective solution to these problems that can be borrowed from an industry with similar issues, and enhanced to be more suitable to this industry. Competitions have emerged within the `Big Data’ community to evaluate analytical capability–large datasets with known properties are distributed to participants, and they are tasked with making inferences from the data. The results are compared automatically, and the results are posted online for anyone to see.
Jerry has been working with a few industry leaders to recast this process for 3D imaging, and a pilot project is nearing completion. Jerry explains, "It is my hope that an ability to evaluate performance required for applications will drive performance standards for instruments and algorithms, and that the methodology when properly adjusted will be applicable on many different scales and across technologies. It remains to be seen whether the industry will embrace the idea."
When asked where he thinks the industry will be in 2020 Jerry predicts, "Clearly 3D reality capture in many forms will continue to explode, and I expect new things to happen very quickly. The most important of them we are not even thinking about at the moment. There is clearly huge pressure to democratize imaging technology and improve the experience–capture, applications, storage, access, COST.
Reflections
When Jerry was asked if he thought from the beginning that the laser scanning industry would prove to be a good place to invest his time he jokingly commented, "I was determined to be successful technically and achieve some level of financial independence. I was blissfully unaware of just how unlikely that might be."
Jerry believes that the following were the key events that have shaped the laser scanning industry over the past 10 to 15 years:
Miniaturization and the consequent improvements in performance and cost of components–optical, mechanical, and electronic.
Software, systems, and algorithms capable of dealing with large sets of points.
Improvements, still in their infancy, in extracting and presenting information from raw data.
Has he ever thought about switching careers? Jerry answers, "Well, yes and no. I didn’t start out thinking I would become a lidar engineer, and I am not sure I need to design another laser scanner. I have always been interested in lots of different things, though when building a company, particularly personally funded, focus is essential. So now I can afford to pursue additional things."
In addition to his technical career Jerry is an outdoor adventure sport enthusiast. Jerry notes, "I try to pedal my bike several times a week, and ride several centuries a year as well as often a longer tour–this year I plan to do Cycle Oregon in September. I have been an avid whitewater kayaker, and have an expired instructor’s license. I try to surf occasionally, despite being the world’s worst surfer."
For his closing comment Jerry borrows a quote from his highly valued partner at Voxis, Joe West–"Experience is what you got when you didn’t get what you wanted."
Gene Roe is the Managing Editor and Co-Founder of LiDAR News.
A 2.228Mb PDF of this article as it appeared in the magazine complete with images is available by clicking HERE