Book text is in a living document format, licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License. In other words, you are free to share this work for non-commercial purposes as long as you don't create derivative works of this material, attribute me as the author, and keep this license notice intact. For more information about this Creative Commons license, please visit: http://CreativeCommons.org.

Chapter 10

simulation for the collaborating mind

One of the more interesting phenomena I've noticed in academia among researchers is how timid specialists can become when venturing outside of their domain of research. When sitting in a group setting with accomplished researchers who have come together to discuss the opportunities afforded by a collaborative grant proposal effort, I've noticed how uncomfortable individual researchers can become when discussing other knowledge bases to be explored in joint work. As soon as the meeting turns to discussion of the subject in which a researcher is expert (the title earned through insightful publications and a strong contribution through admirable investigatory work practices), he or she comes alive with full access to their explanatory and persuasive abilities.

Looking back at chapter 7 and the case study of simulation for understanding natural watershed systems, we can imagine a meeting among scientists discussing the impact of their individual research to the collective study of watershed health. A geologist talks about the effect of the last ice age that carved out the most recent macro-profiles of the terrain and water channels. A freshwater biologist discusses the last six months of water flow and the effect of that flow on fish health. One scientist is thinking in terms of tens of thousands of years' time scale and the other thinking in terms of time scales within a year. Both have valid perspectives on potential human impacts within the watersheds but they feel uncomfortable exploring time scales outside of their day-to-day perspectives within their fields. By putting in years of study and contemplation at their time scale, they have each gained a tremendous appreciation of the significance of time and its impact on their thought process. As a result, they appreciate the effort others have put in a different time scales and realize their thought process is not going to be able to duplicate that focus in the length of a single meeting, let alone a year's worth of research.

For two scientists to collaborate on a single grant opportunity, they need to find a way to respect each other's opinions without necessarily understanding where they come from in their entirety. They need to find a common ground in which to pursue meaningful research. If the common ground is in the middle ground between them, a bridging of knowledge bases needs to take place. A simulation can become a tool that permits those knowledge bases to be explored — especially if the tool can represent both knowledge bases and the expert perspectives of those who work with them. Such a simulation can benefit by being built explicitly to represent knowledge from diverse domains while at the same time seamlessly integrating them into a systemic representation of the phenomena being simulated.

Through the use of the simulation, a collaborator can hone in on other domains and think them through in their own language before becoming subject to the jargon of another collaborator's explanations. They can start with the contribution their own knowledge base provides as an anchor to unexplored domains. Collaborators can comfortably interact with a simulation to probe phenomena at their own pace without worry of looking naïve or perhaps feeling foolish in their lack of trained perspective.

Consider the case study from chapter nine where both trained nurses and trained firefighters worked together to evacuate a hospital. The simulation can provide a trained nurse the opportunity to explore the transport of patients without being an expert in transportation. The simulation can provide a trained fire fighter the opportunity to explore the nature of patients in a hospital setting and how nurses evaluate their readiness for moves elsewhere. That's perhaps an easier bridge to divide since both the trained nurse and trained firefighter are working within the context of human beings of which they are one themselves. We'd expect it to be harder for a geologist to take on the perspective of fish lifecycles or a biologist to be able to think in terms of millions-of-years epochs. The more difficult the bridge to build, the more the simulation (if built with a visual language that represents phenomena well) can potentially assist in the bridging.

Collaboration among people has never been easier when it comes to geographical hurdles. Our information highway, built upon the Internet into a Web of collaboration services, allows any two people to communicate at the speed of light by converting their thoughts into messages that can be transported electronically. The most recent Web services, integrated better and better into Web browsers as we go along, are taking advantage of two-way communications through flexible peer-to-peer networking specifications. The Web exists and supports collaboration so reliably thanks to a fantastic collaboration of knowledge base specialists around the world. They collaborate seamlessly across their knowledge bases as they have learned how to participate in a process of building unique solutions to issues associated with Web architecture. Collaborators have learned how to incorporate solutions within their areas of expertise that interact well with other innovations that then communicate well with theirs. The process of building and maintaining an electronic global collaboration platform is a great example of how to use a platform to collaborate in its own right.

At the extreme ends of the range of necessary knowledge bases that are tapped to make the Web function well, we have physicists who understand the limits and behaviors of physical laws and phenomena, and we have artists and cognitive scientists who understand how to build attractive, intuitive interfaces for humans to collaborate effectively. In between those extremes, programmers connect those interfaces to functioning software. Logistics engineers connect that functioning software to services that can maintain sane and reliable collaboration sessions among the millions who are using the Web at any one time. Network engineers reliably makes sure the content of those sessions makes it from start to end points among the collaborators. Those physicists advise the development of devices that can generate useful physical phenomena that can carry meaningful signals across reliable communication channels (along wire, fiber-optic cables, through the air, etc.). Hardware engineers build the many different types of devices that the physicists help design as feasible and practical to take advantage of physical laws. Chemists help with the materials used in creating reliable, cost-effective, and safe devices.

All of these experts of necessary knowledge bases associated with the Web work together thanks to the overall structure of the Web — without needing to meet each other or absorb the depth of understanding of each other's expertise. The Web provides specifications and protocols that allow anyone to add a device, communications channel, or new application to the overall Web enterprise at any time. As long as those who do so are cognizant of the potential disruption their technology may provide, they are encouraged by the community to try out their ideas over short time periods to test out their invention. The Web is built to be able to function well even if a part of the Web suffers negative effects from some new phenomenon. Instead of simulating the effect of new components on the overall Web, we can try them out and then interpolate the potential effect as we scale up over time and distance. It's as if the existing Web is a simulation at any time of what would happen if the Web were configured as it is at that instance. And yet, there are simulations of the Web that people use to anticipate the effects of different Web uses. Much of the time they are built to consider a development project without having to actually build a prototype first.

The secret to the great functioning of the Web over time is the hierarchical structure of the knowledge bases involved. The physicist's expertise can be applied to the behavior of devices and physical communication media without any concern for the logistics engineer, network engineer, programmer, artist, or cognitive scientist. The artist's expertise can be applied to the design of human interfaces without concern for any other knowledge base but the programmers. And, yet, there should be such a great appreciation of all the knowledge bases as the Web would not function so usefully, reliably or efficiently without all knowledge bases combined.

We crave that same sense of collaboration to gain insights when we are taking a systemic view of any complex phenomenon. Not everything is hierarchical in nature so we might not be able to make such a clean suggestion of which relationships among knowledge bases need to be tightly coupled. But, certainly many things in nature are hierarchical and we can start there for the rapid progress that is possible. Those who study human perception and cognition have developed hierarchical models of brain functioning that seem to predict behavior of the brain quite well (the visual system, for example). Is the process truly hierarchical because hierarchy is a structural advantage evolution has taken advantage of? Or, have we used a hierarchical understanding because we have evolved thought processes that progress better when we force phenomena into hierarchical structure whether they are actually hierarchical or not?

Simulations can take advantage of the Web to run among collaborators from afar. There are some powerful examples of that concept running in labs around the world. Consider the following two examples:

The O.H. Hinsdale Wave Research Laboratory at Oregon State University maintains a physical wave tank that can simulate repetitive wave forces on beach, land, and building structures. Since the original tank in 1972, the physical laboratory devices were updated to focus on three-dimensional coastal issues and tsunami research within the network for earthquake engineering simulation. Collaborators can present virtual building designs that have been tested in virtual wave simulations to be tested as physical structures within the physical wave tank. Virtual tsunami models can provide wave amplitude and frequency outputs that can be generated as physical waves at a specific reduced scale. Projects scale has been as large as 1/3 of actual size for small physical arrays of devices. The wave tank then tests the structure of coastal buildings for possible effects of tsunami wave action through collaboration with building designers via the Web.

Robotics projects often begin with simulated robots in order to incorporate the knowledge bases of multiple domain experts. One of the more popular examples is Sojourner, a mobile robot that currently operates on the Martian surface. Before Sojourner could become a functioning physical robot, collaborators had to design solutions that met the needs for the various systems that needed to work reliably in the design of a single coordinated robot. Telecommunications, mechanical, thermal, mobility, control, navigation, and power were important sub-systems developed by experts to perform the activities set out by a NASA mission to Mars. Simulations are still used to develop solutions for problems or opportunities that are experienced by the rover on Mars. Terrain models of Mars are used to anticipate time and motion needs for the rover through simulated navigation before attempting a physical route on Mars.

Both the wave tank and Martian rover represent successful projects that bring together disparate knowledge bases into functional virtual simulations and functional physical devices that provide great service to humans through science research. Those projects are costly projects performed with high stakes implementations by experts who already have a track record of demonstrated expertise to their names. The simulaton life suggests all of us should have the opportunity to participate in low cost projects performed with low stakes implementations that can be iterated upon as expertise is developed among collaborators. Simulation development skills are a form of knowledge acquisition similar to reading, writing, mathematics, and geography. We can develop skill in creating, editing, using, and verifying simulations as we gain insight and understanding of phenomenon being simulated.

Whether hierarchical or not, our simulations can be decomposed into modules that tap knowledge bases for the effect each module has on the whole. We have the opportunity to build modules in ways that many simulations can adapt them for their uses. By aligning modules along the lines of expertise, we let those experts collaborate and make progress on their own time and effort without relying on others to keep up. As they can describe the use of their module with other modules theirs likely effect most dramatically, they can build interfaces with other groups of experts working on their own modules — expressing their own knowledge bases in contributions to the overall simulation.

There are those among us who thrive at looking at things across knowledge bases. We readily find patterns of interconnection that can be reused to connect knowledge bases together in meaningful ways. Some of us with that ability have been formally trained as systems thinkers or systems engineers. Some of us have never been able to sustain our full passion going deeper past a certain point of understanding in knowledge pursuits as we naturally pursue knowledge breadth instead of knowledge depth. How many of us are out there? It's an interesting question and one that likely sheds light on the true value of supporting a simulaton life through experiences with simulation tools. I hope I have done the right things to become better at looking at things across knowledge bases. I feel like my best days in my life have come when I find a new thought that interacts well with the models I have developed of the world thanks to previous thoughts and experiences. That new thought seems to ripple through my memory and play with what I have remembered. As a result, I feel my brain generating insight after insight and on the best of them I find my brain has reconfigured itself to have a significant new perspective towards the world — one that changes my behavior or at least builds conviction in my intent to change behavior.

Our academic institutions are not usually built to support collaboration first and foremost. Universities reward faculty on a tenure track for providing service to a department through tasks of teaching, researching, and writing. The department has a name that defines a particular knowledge base — a chemistry department or accounting department. Members of the department can best evaluate other members when they understand their jargon and their domain. They can evaluate students as students attempt to learn the jargon and the purpose for such efficient use of words. Collaboration may happen between research institutions through common interests but the bonds between faculty members are strongest among those who share the same expert knowledge base, but not necessarily the same opinions about the value and truth of the components of that knowledge. Evaluation is important in education and so the status quo remains for the benefit of assessment. Collaborations are more difficult to assess without a well-defined mission like a successful mission to Mars or saving lives when the next tsunami or community crisis comes along.

Simulations can be used for assessment when the simulation attempts to simulate real world phenomena. The assessment comes from the comparison of the simulation to the actual physical phenomenon. Although weather forecasting sits squarely within the meteorology department at a university, the simulations weather forecasting drives can span across university departments. A good watershed simulation requires expertise among many specialties in the earth sciences. Watershed simulations can be assessed through a ground-truthing process that compares simulated results to observed results in the physical world. As the simulated results diverge from physical results, new insights and understanding that may explain the differences feed back into specific knowledge bases of those involved in the collaboration.