Edited Version of May 12, 1998 Transcript
EIIP Virtual Library Online Presentation
"An Interactive, Intelligent, Spatial Information System "
Dr. Louise Comfort
University Center for Social and Urban Research
University of Pittsburgh
The original transcript of the May12, 1998 online Virtual Library presentation is available on the EIIP Virtual Forum (http://www.emforum.org). The following version of the transcript has been edited for easier reading and comprehension. Typos were corrected, date/time/names attributed by the software were deleted but content of discussion, questions, and responses are as stated by each participant. Answers from the presenter to questions by the audience are grouped beneath the appropriate question to facilitate meaning. Related questions and discussion from the Brown Bag session (immediately following the presentation) are included in the edited transcript.
Amy Sebring: On behalf of the EIIP, I am pleased to welcome you to a special event in our library. Because of changes to our software, we are having an UNMODERATED session today. It is my pleasure to introduce Dr. Louise Comfort, Associate Professor, University Center for Social and Urban Research, University of Pittsburgh. Dr. Comfort is also the Principal Researcher for the IISIS project which she will describe for us today.
Louise Comfort: Thanks Amy. I am glad I could make it here today to talk about "An Interactive, Intelligent, Spatial Information System (IISIS): A Community Model".
The problem: the costs of disaster have increased dramatically, estimated at $50 billion a year in the years of heaviest losses, 1993 - 1995. Stark increase in costs of disaster reflects growing interdependence of metropolitan regions and a society with an increasing population and advanced technologies.
Communities function with increasingly interdependent systems -- transportation, communications, power, water, sewage, natural gas. Major causes of failure: lack of knowledge of the consequences of failure of one system upon others; lack of timely, accurate, information to mobilize effective action.
Three basic issues are involved:
1. Practical: how can a community increase its capacity to mitigate and respond to disaster in its environment?
2. Organizational: how do experienced personnel make decisions in response operations under rapidly changing conditions?
3. Technical: how can an information system support the changing roles of personnel and organizations as they adapt to shifting demands from the disaster environment?
Unmet needs: Existing software programs designed for emergency management have not successfully met the needs of practicing managers.
We suggest four main reasons:
1) lack of flexibility in representing the diverse perspectives in a dynamic situation
2) lack of staff training and current data in the knowledge base
3) lack of inter-organizational decision support that is crucial in a community-wide event
4) lack of an intelligent reasoning component that can assist managers in addressing the rapid increase in volume and complexity of information in disaster events.
The IISIS Approach:
1) Links information technology with organizational design and learning to provide decision support to practicing agencies
2) Creates a 'sociotechnical' system of interacting computers, organizations, and individuals working toward a common goal of reducing risk.
3) Supports a dual use system for improved management of disaster at both organizational and community levels.
4) Long-term goal: to increase the capacity of communities to mitigate and respond effectively to disasters.
Design issues for an information system to be used in disaster environments. Such a system:
1) addresses information needs for decision-makers operating in the dynamic, urgent environment of disaster operations
2) needs to be more flexible than a typical information system to accommodate novel situations and new roles
3) provides not only a knowledge base of relevant information and a record of unfolding events, but also an intelligent guidance system to support individual reasoning and cooperative decision-making
4) needs to be dual-use, so that staff would use it in their daily operations and update the knowledge on a continuing basis.
The current IISIS prototype consists of three basic components:
1) interactive communication among distributed knowledge bases
2) dynamic maps as the event unfolds, using GIS; and
3) intelligent reasoning by the computer to integrate incoming information from the environment with stored information about the community.
Figure 1 shows how observations of a threat are transmitted via a Web page interface to an information system that provides a continuous assessment of the 'state of the community'. First slide please,
Louise Comfort: As new information enters the system, it is passed to the intelligent reasoning component, which compares the change reported about the community with stored information from the knowledge base to assess the difference and estimate the degree of risk to the community.
The human-computer interface is a set of Web pages that receive input from the user, transmit it to the information system, and return new information, processed through the system, to the user.
The set of functions for the IISIS prototype is organized in three basic subsets. The first identifies the existing conditions in the community as the basic knowledge base for coordinated action. The second subset reports the event and assesses its impact upon the community dynamically. The third subset provides a continuous record of disaster operations for the specific disaster event.
Figure 2 shows the field status screen, which allows the user to enter the incident report: time, location, type of hazard, estimated severity, environmental conditions such as temperature and wind direction. Next slide please, Amy.
Louise Comfort: The next three figures show aspects of the dynamic assessment of the event. Figure 3 shows the Disaster Site, Community Scale. Next slide please, Amy.
Louise Comfort: Figure 4 shows the Disaster Site, building scale. Next slide please, Amy.
Louise Comfort: Figure 5 shows the Damage Assessment screen. Next slide please, Amy.
Louise Comfort: The intelligent reasoning component will provide continuing assessment as the event evolves. It will identify the area of geographic impact, and within this area, estimate the consequences of this event for the population of the community, its built environment, its lifelines and existing resources for response.
This information is posted on the 'electronic blackboard' which is monitored by the knowledge sources for the separate organizations that have legal responsibilities for response in disaster: fire, police, emergency medical services, public works, and others.
The organizational knowledge sources contain the standard operating procedures for each organization in event of emergency. If any one of these procedures is triggered for action by the incoming information, that action is reported to all other emergency response organizations via the electronic blackboard.
Louise Comfort: The intelligent component uses this information to calculate the probability of risk to individual sectors of the community, as well as the cumulative level of risk for the whole community. This information is returned to the user (practicing manager) by the computer via the Web page.
The third set of functions provides a continuous record of disaster operations for the current disaster event. Figure 8 shows the current status of operations for the event. Next slide please, Amy.
Louise Comfort: The prototype IISIS tracks the pattern of interactions among participating organizations, creating a common base of knowledge for the event and facilitating mutual coordination of actions among the participating organizations.
Louise Comfort: The IISIS prototype adapts and integrates appropriate information technologies to support interorganizational decision making and providing real time information to practicing managers. The use of the IISIS prototype will contribute to a community's capacity to adapt and respond to demands from its environment in a more timely, appropriate, and efficient manner. I hope this conveys the essence of the paper. I would be happy to take questions or comments now, Amy.
Amy Sebring: Thanks Louise. I would like to take a moment here to review how we will handle the Q&A so that we have an orderly session. We ask that you indicate that you have a question by typing just a question mark (?). Then you can prepare your question, but PLEASE HOLD (don't hit end or send) your question until you are recognized. If we run out of time, you will have a chance to ask Louise afterward in the follow up session in the Virtual Forum. Ready for questions.
David Crews: How do you propose to cross the traditional political and bureaucratic boundaries who keep data that is proprietary?
Louise Comfort: David, this is the hardest part. The main issues in using information systems is not the technology. It's getting organizations to work together, to share data and to accept a common set of standards.
Kevin Farrell: Louise, I'm assuming that you've tested the model, but what sort of situations (large and small) have you tested it with?
Louise Comfort: We have not actually used the prototype in a real situation. We've used it for training and to allow organizations to discover they need to share data.
Kevin Farrell: Simulations?
Louise Comfort: Yes, primarily simulations.
Rick Tobin: One of the toughest parts of Artificial Intelligence systems is input and maintenance of relevant data. Any strategies on getting those kind of commitments in a world of bottom line management?
Louise Comfort: It has to be a dual use system. That way, the personnel who use the data every day are updating it constantly. People will discover their efficiency increases with better information on daily use also!
Rick Tobin: I can't imagine most managers using this on a daily basis. At least, not yet.
Louise Comfort: Agency staff need to communicate with their personnel. It helps to do it on a regular basis.
Amy Sebring: Louise, can you tell us more about the reasoning component, what type of engine you are using?
Louise Comfort: We are using a blackboard system that was developed at SRI. Bruce Buchanan who worked on the system is now at the University.
Amy Sebring: Is that a rules-based system?
Louise Comfort: Yes, which fits the sop's that emergency managers must use. They understand clearly the parameters within which the machine is reasoning.
Amy Sebring: SRI Louise? Is that research or private?
Louise Comfort: Stanford Research Institute, mostly private, but with lots of federal grants. The software, paid for by tax dollars, is now free.
David Crews: Have you worked with other universities on this? I received a call from University of North Texas interested in a info/comm partnership with the business community and in forming a core group to determine the need for a strategic plan.
Louise Comfort: I've been in contact with most researchers working on the problem. No one else that I know is working on an intelligent reasoning component.
Joanne McGlown: Louise, I am assuming there is a base of data resident, continuously. If correct, what type? Demographics? GIS mapping? or if off-base. (sorry, couldn't see the slides well)
Louise Comfort: We are seeking to combine many databases for the region. The prototype will run best on a well-designed knowledge base. GIS will be a major component of this system.
David Crews: North Texas U wants to do something like this.
Rick Tobin: I believe California State OES would be very interested in the process you are developing. Is there a formal paper or presentation available on this? We've been discussing AI for 2 years here, but nobody has attacked it as you have. It could be a marriage made in heaven.
Louise Comfort: Rick, I would be delighted to work with Cal OES. I'll be at Berkeley this summer and would be very glad to meet with you. In fact, I'm leaving for Berkeley tomorrow, just for the weekend, to attend a NSF conference on EQ hazards.
Amy Sebring: Louise, in follow up to Joannes question, are you following the FGDC concept of core data elements?
Louise Comfort: Joanne, the FGDC (Federal Geographic Data Committee) concepts are very important. It really outlines the design for a national information infrastructure, which we badly need. The FGDC has developed standards, and if all communities would follow them. I think it will be solved, but it will be time-consuming and expensive.
Joanne McGlown: Louise, I also must raise the Y2K issue when talking about compatibilities. Do you feel this will be an issue for this prototype?
Louise Comfort: Joanne, it will be an issue only in integrating different kinds of data but there are major efforts underway to design a conversion program.
Amy Sebring: How far along are you with the prototype, and how far to go?
Louise Comfort: We have the system design nicely developed. We have the Web pages and a database for one jurisdiction reasonably developed. We are still working on the intelligent reasoning component which is the hardest part. We would very much like to create a consortium of public agencies, universities and private companies to work on this problem. Yes, EIIP is the model!
Avagene Moore EIIP: With whom will you do your field testing? Will it be done in stages or how?
Louise Comfort: For licensing reasons, we have been using the University of Pittsburgh as a field site. If we use it only at the University, our costs are much lower. But we would like to work with practicing agencies for a real test.
Amy Sebring: The University is a microcosm of the broader community, is it not?
Louise Comfort: The University is actually a third-class city of some 32,000 souls. It has a hazmat team, campus police, emergency medicine and lots of unruly people.
Amy Sebring: What is your approach to gathering the rules for the AI component Louise?
Louise Comfort: We've used several sources. First, are the emergency plans and the legislation, policies governing disaster management. But more important, is the set of field observations I have done over the past 15 years and crucial are the interviews with practicing managers. It is important, however, to follow them around, because they often do very innovative things which aren't in the rules and which they don't think to tell you.
Amy Sebring: Or to write down anywhere!
David Crews: North Texas U approached me last week wanting to build a Info/comm consortium and have already talked to the States EM in California, Florida and Texas. It sounds like the academic community has several parallel efforts going. As a suggestion maybe a consortium in the academic community would be the place to start, using the IISIS as part of the foundation.
Louise Comfort: I am aware of the people at North Texas, I think that universities could be a major resource in building the information base.
Avagene Moore EIIP: Louise, you mentioned that we need a national information infrastructure; from your perspective, where are we in such an effort and what do you see on the horizon?
Louise Comfort: We are behind. People are just beginning to be aware of the need. We need major public funding, shared with private and nonprofit organizations to build the knowledge base and to train personnel to use these systems. I would like to see universities as a major leader in the field. They can do the research, test in partnership with local agencies and then work with businesses to produce and market the software for public use. It would save billions in losses.
David Crews: NTU is proposing major funding from private sources instead of relying on public funding.
Louise Comfort: We may need to go that route but I worry. Emergency management shouldn't have to depend on the market.
Amy Sebring: We can follow up Private vs. Public afterward in Virtual Forum.
Amy Sebring: Louise, have you considered adding environmental sensors as a component in the future?
Louise Comfort: Yes, we will be doing that in a prototype we are designing for the Czech Republic.
Amy Sebring: Can you tell us just a little about that?
Louise Comfort: This is a major project. Still just in the planning stages. But the Czech Republic has a severe problem with environmental pollution. In order to get their economy functioning, they need to clean up the mess from the last 50 years. We will be working with a consortium of public, private and nonprofit agencies, with funding -- hopefully -- from the World Bank. The University of Pittsburgh will work with Czech colleagues in setting up the design and training their staff.
Amy Sebring: Thank you very much Louise for sharing your work with us today. Thank you audience, and since our time is up, we will close down the Library for today, but we will be in the Virtual Forum room for a few minutes longer, and you are welcome to join us there for open discussion. Thank you for your cooperation.
Further Discussions and Questions in the Virtual Room - After additional expressions of appreciation to Dr. Louise Comfort, a few participants stayed for a few minutes with additional comments and discussion related to the library presentation; the following are excerpts that convey additional information about " An Interactive, Intelligent, Spatial Information System, (IISIS): A Community Model".
Amy Sebring: I am with Louise on public funding, I don't think there is any money in this.
Louise Comfort: I think public funding is the only way. The national information infrastructure will be as important to the future of this country as the railroads. But we can certainly include businesses and non-profits in the effort.
Joanne McGlown: Great job, Louise. This is fascinating work. I look forward to hearing more about it.
Amy Sebring: Thanks to all. Don't forget the transcript will be available as well as the slides by this weekend.
Louise Comfort: Thanks.
David Crews: Your effort certainly demonstrates the potential of the Academic Community in developing Research, Development, and Policy for EM.
Louise Comfort: I think we have something to contribute; bright young people, curious minds, the next leaders of our cities and states.
David Crews: Academia can also cross some of the political and bureaucratic boundaries that hinder development of such systems in Government and Business.
Louise Comfort: Yes, I think we can. Universities can't make any money. But we can distribute ideas.
Avagene Moore EIIP: We are trying to tie academia and other emergency management interests more closely together. Badly needed and I think the EIIP can help.
Louise Comfort: These projects are very expensive. And do require significant funding for a period of years.
Avagene Moore EIIP: Yes, and funding becomes a more serious issue every year.
David Crews: Some sort of joint funding mechanism needs to be found that is reliable and has longevity.
Louise Comfort: David, I wholly agree. And this is where EIIP can help.
Louise Comfort: If there are no more questions, I must get back to work. Thanks!
David Crews: Thank you!
Avagene Moore EIIP: Thank you, Louise. Appreciate you and your support of the EIIP and the emergency management community. Well, folks, anything else to be said or any other questions. Join us tomorrow for Wayne Blanchard's session. Will be good and enlightening. The sessions with the folks in Puerto Rico on Friday and Saturday will be good too. Thanks for being here.