Sigma Xi 125th Anniversary Interview

Ed Mikhail (SX 1961) 
Interviewed by Melissa Rura (SX 2010)

What made you want to be an engineer?

For as long as I can recall, even when I was a small boy, I always wanted to be an engineer, and specifically a civil engineer. One time, when returning from my Franciscan elementary school in my childhood hometown of Fayoum, Egypt, I stopped by a construction site. There buildings are constructed with concrete columns and floors, and brick walls almost exclusively. I was curious why steel bars, which I later found out was rebar, were used. I recall asking the nun teacher that question, to which she replied that in order to understand that you will have to become an engineer. She said “Edward, you are very good at arithmetic, so if you study hard you can be the first in your family to go to the University and study engineering.” This turned out to be a prophetic statement; my three older brothers did not advance past secondary school. Not only did these dedicated and often strict, sometimes harsh, nuns set me on a path to a good future, they all instilled in me an excellent work ethic and the principle that you would reach your goals with unwavering commitment and hardwork. My childhood curiosity lead me to propose and design a composite steel beam / concrete deck acting as a unit, for my BS project, after which I continued on to a M.S. and Ph.D.

What has been your most fulfilling accomplishment as an engineer? Why?

After my B.S.C.E., I received an appointment as instructor at Cairo University and worked as a structural engineer in the evenings at the largest consulting firm in the Middle East. During my graduate studies at Cornell University, I spent my summers as an intern with different United States and Canadian governmental organizations. Thus, by the time I finished my Ph. D., I had experience working in three job sectors: education, industry, and government. I elected to pursue a career as an educator because helping students understand complicated concepts was rewarding. I was fortunate to join Purdue University in 1965 and establish a graduate program in my specialty, photogrammetry that same year. At Purdue I was able to teach, perform research, and also consult, a very rewarding career. It is not the hundreds of papers and reports; not the books I authored, co-authored, or edited; not the patent; but the pleasure of helping several hundred M.S., and several dozen Ph.D. students matriculate that is by far the most fulfilling accomplishment. Why? Because it is by far the most enduring. I consider most of my students my friends. I have watched many of them flourish into excellent careers and humbly accept a very small measure of satisfaction at having been lucky enough to have had the opportunity to help them, a little, to attain their goals. I recollect a gentleman joining our program at an age many years after his college graduation. He had been in practice, and found it exceedingly difficult to return to “school life,” not to mention having a tough time with recollecting his early math and physics concepts. He all but quit half-way through his first semester. With patience, encouragement, and working with him one-on one, not only did he finish, but he became an excellent educator, winning teaching awards. When he was promoted to full professor, he sent me a very touching letter. What can be more fulfilling than helping others?

What significant changes have you seen in your field during your career?

Originally, I was deeply interested in structural engineering and could not imagine myself working in any other field. However, due to factors totally beyond my control, photogrammetry became my specialty during my graduate work. Even then, I enrolled in as many structures and soils mechanics, now geotechnical engineering, courses at Cornell, fully expecting to consult in these first love areas, while teaching photogrammetry at Cairo University. Needless to say, photogrammetry captured my interest and I ended up fully involved in it and its development. Perhaps because it was, at the time I joined Purdue in 1965, a relatively young field. It has since seen enormous changes. At a very high level, it is impacted by three factors: sensing systems, platforms, and processing. When I started, film-based cameras were essentially the only sensing system. At present, sensing has expanded enormously -- beyond the visual range of the electromagnetic spectrum — to include radar, sonar, laser, among other wavelengths at multi and hyperspectral bandwidths. Sensing also moved from the simple central prospective of an analog camera to various scanning schemes, while light-sensitive emulsion film has all but been replaced by linear and square arrays of charge coupled devices (CCDs). The addition of satellites to airborne and hand-held platforms has significantly expanded the fields of photogrammetric application by making more accessible vast areas of the earth at different scales of collection. Finally, the explosive development of computational processing has opened a new era for photogrammetry. The first phase of photogrammetry was termed “Analog,” where optical and mechanical components simulated the imaging geometry. I began my career at the dawn of the “Analytical” phase, where mathematical modeling and rigorous estimation techniques from redundant data were developed and implemented. The third phase is “Digital” photogrammetry enabled by both digital image data, and fast digital processing. I believe the next phase will likely be what I term “On-Demand” photogrammetry, facilitated by web services and fast at near real-time processing. The fundamental concept of photogrammetry is that a 3D object can be faithfully and accurately recovered from two different images, usually termed stereo photogrammetry. I recall in 1992 serving on a National Research Council (NRC), (National Academies) Committee on Strategic Technologies for the Army of the 21st Century (STAR-21) Long Term (i.e., 25 year) Forecast of Research. Perhaps too boldly, I predicted that there “will be” a system in the future that would yield a direct 3D image and there would be no need for “stereo” reconstruction. Back then, nearly twenty years ago now, no one would have imagined what we have today: complete systems of 3D surface mapping using different types of laser scanning modes. Further, I read very recently that a “3D camera” has been developed. I guess there is no harm in being “bold” when it comes to scientific advances.

What would you consider to be the most important advice you could offer a younger, upcoming scientist / engineer?

I am going to assume that he/she is in a field in which they completely enjoy working. There is nothing better in this world than getting up every morning looking forward to “going to work.” It has been a blessing for me for over 50 years. The opposite is of course rather sad. Under this assumption, then, my best advice is: NEVER lose your curiosity and BE inquisitive. Keep in mind when, from your curiosity and questioning, ideas come, to not let anyone tell you that “such a thing” is impossible. History is full of stories of people being dead wrong about things they said cannot be done. One of those, who is considered “one of the fathers of photogrammetry,” had said in his analog time, that the formulation of a rigorous camera model was so complex that it would never be possible to numerically solve the resulting system of equations!! Formulations of current sensor systems are far more complex, but then, the computational tools are more than capable of handling them. Another piece of advice for a young scientist is to broaden his/her horizons. Some of the best opportunities come in collaborative fields. For me it came by being intrigued by optical holograms in the 1960’s, which motivated a cross-disciplinary research effort leading to the “holographic stereogram,” which was a most gratifying accomplishment.

What do you think are the most pressing needs to be met in science in the coming years? What would you like to see scientific research accomplish?

To me, the primary goal of science and engineering should be the betterment of people’s lives. There are several facets to accomplish this goal. First, “resources” – that is food, water, and energy. There must be concerted and innovative efforts to alleviate hunger, including improved quantity and quality of food supply and distribution, water management and recycling and finally a strong and deliberate push for all types of renewable energy sources. Second, “health” – nothing can have a more positive impact on humans, than not only finding cures to tough illnesses, such as various cancers, but perhaps better yet, finding preventative means of healthcare. Third, “infrastructure”—effective and economical means of upgrading current infrastructure, and developing newer material and intelligent transport and communication systems. Fourth, “exploration”—for natural resources exploiting advances in remote sensing; of space; and of the seas so as to markedly expand the human frontiers, and to continue pushing forward the exploration of space.

As part of Sigma Xi 125th celebration, we are focusing on ethics and responsible research. Have you seen any changes in ethical conduct within your field during your career?

Only on rather rare occasions did I encounter changes in ethical conduct. At least, not anything comparable to what I have read of the changes in other fields. Ethics constitutes the pillars upon which research rests. The public must have absolute faith in the integrity of the researchers preparing research results. Any action that may shake that faith must be dealt with, steadfastly. Educators and supervisors of research have to be quite diligent and rigorous in imparting to students, juniors researchers, and colleagues the sense of absolute integrity in carrying out experiments and properly analyzing results, so as to make sure that ethics are never compromised, regardless of any disappointing results.

One of sigma Xi’s strength’s is its interdisciplinary nature. How important do you think interdisciplinary collaboration will be for solving some of the challenges that lie ahead in science?

Quite important. As I mentioned in an answer to an earlier question, one of the most valuable discoveries in my own career was the result of a collaboration research project between my area in the School of Civil Engineering at Purdue and that of the coherent optics group in the School of Mechanical Engineering. Another was a five year project that I initiated which included: my area as lead, Electrical Engineering at Purdue (Remote Sensing and Pattern Recognition), Electrical Engineering at University of Southern California (Computer Vision), and a large corporation in San Diego, CA (Visualization); a most gratifying project!

I firmly believe that the myriad of problems in the future will most likely be solved by teams of scientists / engineers from different disciplines bringing their diverse expertise to seek novel and effective approaches. We have already seen how advances in, say, sensing and automated recognition have significantly affected developments in transportation, for instance, by introducing the so-called intelligent systems. Geospatial Science and Engineering has pulled together diverse areas of activities and promises to play a pivotal role in solving challenges in planning, exploration, and resource management to improve the human condition.

Where would you like to see Sigma Xi in 125 years?

It is hard enough attempting to predict 10 or 20 years out; 125 years is far beyond any man’s ability … certainly mine! So, the best I can do is to venture a few ideas:

  • The inclusion of successful (i.e., ethical and inquisitive) future scientists / engineers should start early in their secondary education (I am one humble example); perhaps launching an effort to foster love of scientific exploration in , say, high school, by working with the teachers… or even “junior clubs.” A high level periodical that can be comprehended by high schoolers maybe worth publishing. I expect that all this may be best facilitated on-line through the web.
  • In the spirit of “cooperation not competition,” establish a “submit a question” service. A researcher may encounter a particularly difficult problem, which may take a considerable amount of time for him/her to solve by themselves. Submitting it may very readily be within the domain or expertise of someone who could quickly provide, if not the complete answer, at least a valuable input or direction as to lead the researchers to a much faster resolution.  This service would be like a “global research advisor,” in the role of the dissertation advisor, significantly increases the “globalization” of Sigma Xi. I can visualize having “affiliates” in many other countries that are currently “developing.” Often talented people who reside in developing countries are much closer to the human needs and problems around the world and by facilitating their connectedness to the larger scientific community both developed and developing countries are more likely to propose and contribute more effective research solutions. 
  • Establish a “global on-demand library” with the most up-to-date research results and current science news. There is no doubt that near, if not, real time communications will be commonplace. It will help improve the general knowledge about science to every person worldwide. 

What is the biggest open question in photogrammetry that will require the most attention in the future?

Photogrammetry provides reliable information about objects and environments using imagery and other sensed data. Currently, data processing and data accuracy assessment / quality control that occur subsequent to data acquisitions. Whereas significant advances have been made to increase the robustness of data processing / handling, quality control and data reliability of the subsequent extracted information, and a reduction of elapsed time between acquisition and production of information, there remain at least three major focus areas which will require substantial effort in the future:

  • Reliable (i.e., quality) and fast (i.e., automated) extraction of features from imagery and other supporting data is and has been an open and challenging research objective within photogrammetry for decades. My belief is that a major push to solve this problem will be one of the dominant research tasks for both the near term, working with current imagery and data, and the long term when more sensors and platforms are integrated into an advanced system.
  • “Integrated” data is a second important focus area of photogrammetric research. That is, “merging” or “combining” a variety of data sources (i.e., acquisition platform, data qualities and / or data scales and resolutions) into a cohesive and coherent system, which will improve, facilitate, or otherwise provide value added information to imagery and other supporting data. At present, this task is termed “fusion”, and often its complexity increases with an increase in the diversity of data type, with change in resolution, and with the quality of the constituent data components. Providing a robust (i.e., consistent regardless of data diversity) and reliable (i.e., quality) solution to this problem will be a major accomplishment in photogrammetry increasing its value immeasurably to a variety of application fields.
  • At the present time, much processing of acquired data into information is often preformed downstream (i.e., subsequent to initial data acquisition, initial pre-processing where much photogrammetric adjustment and quality control is done). In my opinion, the most valuable future breakthrough will be shifting most, if not all, of the processing upstream. It would be the ultimate dream come true if in the future, given defined requirements, one can “dial in” a requirement into a “system” which determines the optimum data to be acquired for a particular subsequent post processing goal, acquire that data, process it, and “download” the required pre and post processing “information,” all essentially in real time. Initially the construction and implementation of such a system would incorporate human expert knowledge to design its components and make some support decisions which will ultimately be automated. When fully developed, the role of the human will then be shifted to supervisory role. This means that the human will have the ultimate task of accepting or rejecting the results. Such a system would be immensely valuable for a myriad of very critical applications such as:
    • Monitoring – subsidence, disasters (floods, fires, etc.)
    • Tracking – vehicles, people, phenomena, etc. for national security and otherwise.
    • Acquisition of time sensitive information,

Note that, although I have not explicitly stated it, all such information would be spatially three dimensional in nature, and could eventually be extended to four dimensions by including time in order to allow temporal analysis.