between processes;
CLARIFICATION OF INTER- ABJD IWIRA-
ORGAMZATI O U L O R G A r n r n
1
PROH OTIhlGI
IlfPLEtENTAl'l ON OF DECISIONS
EVENTS WHICH H A Y LEAD
u
Figure 4.1: Motivations a n d perceived efpects i n the u s e o j m . (cfler Humphreys et aL., (2983)).
Since t h e u s e r s , however, are not expected t o b e computer e x p e r t s , a n operational prototype i s essential t o show what can b e done and how. Only on t h e basis of t h e prototype's elements can t h e u s e r s then develop a
sufficient understanding of t h e system's working to t a k e a n a c t i v e role in its f u r t h e r design and implementation.
4.1 A Tentative List of Problem Areas
There are many specific problem areas t h a t c a n b e addressed with t h e system described above. A p a r t i a l and by no means exhaustive list might include:
estimation of waste streams originating from specific industrial pro- duction p r o c e s s e s (e.g., chlorination of phenols);
identification of process modification requirements (e.g., recycling, waste reduction, volume reduction) subject t o waste output constraints (regulations);
design of alternative structures f o r production and waste manage- ment systems (technological, spatial, economic);
exploration of siting alternatives f o r production plants o r treatment and/or disposal facilities, given socio-economic as w e l l as environmen- tal objectives and regulatory constraints;
estimation of trade-offs between alternative production and treatment o r disposal schemes and t h e i r implicit transportation requirements;
analysis of a l t e r n a t i v e regulatory policy options (relative cost and effectiveness) f o r production, use, transportation, and treatment and disposal;
risk assessment for given production facilities or production processes in a specific regional environment;
simulation and evaluation of emergency plans f o r various t y p e s of accidents under a wide variety of meteorological conditions;
risk/cost analysis for t h e transportation of hazardous materials, considering transportation mode and r o u t e alternatives, public expo- s u r e , environmental damage potential, applicable regulations, etc.;
identification of least c o s t / r i s k treatment and disposal alternatives f o r given
waste streams
(amount, composition, t r a n s p o r t a t i o n r e q u i r e - ments);estimation of environmental and public health consequences of various emission scenarios (routine emissions from industrial production p r o c e s s e s o r dispersive use, e.g., agrochemicals, t o atmosphere o r w a t e r , emission from
waste
t r e a t m e n t and disposal, e.g., leaching from dumpsites); such emission s c e n a r i o s might b e d i r e c t l y user-generated o r r e s u l t from any of t h e above applications;long-term simulation of i n t e g r a t e d subsystems f o r t h e description of complete lifecycles of substances (e.g., industrial production,
treat-
ment and disposal, environment) t o identify potential problem areas e.g., disposal c a p a c i t y c o n s t r a i n t s , o r toxics' accumulations above t h r e s h o l d s in environmental media;estimation of environmental hazard ( a v e r a g e a n d maximum ambient concentrations, accumulation in t h e food chain, human e x p o s u r e ) , f o r c e r t a i n s u b s t a n c e s in specific o r g e n e r i c environmental systems.
Many of t h e s e applications will r e q u i r e t h e linking of s e v e r a l of t h e component models of t h e system (compare Figure 3.3). Clearly, although many of t h e impact-related elements in t h e subsequent analysis are t h e same, t h e r e i s a n obvious dichotomy between accident- and waste- r e l a t e d problems.
5. JMPLICATIONS
AND
PROBLEMSE x p e r t systems, dedicated p e r s o n a l computers and professional works- tations are r e l a t i v e l y young phenomena, d i r e c t l y coupled t o t h e r a p i d development of computer a n d communications technology o v e r t h e last decade. A s with t h e introduction of a l l new technology, i t not only solves some old problems, i t a l s o creates i t s s h a r e of new ones.
5.1 The Economic Potential
E x p e r t systems hold promise of g r e a t economic and social impact.
They promise t o b e profitable, because t h e y c a n solve problems t h a t r e q u i r e t h e b e s t and most expensive human e x p e r t i s e . In some domains, t h e exhaustive n a t u r e of problem solving in e x p e r t systems will a s s u r e t h a t even remote possibilities are not overlooked. Obviously, depending on t h e application domain, t h i s may b e important. Also, t h e i r ability t o potentially draw on v e r y l a r g e factual d a t a bases (e.g., on chemicals: ECDIN, developed at t h e JRC, I s p r a , holds information on about 100,000 substances), and
to
t r a c e complex consequences in simulation models by far exceed t h e p e r f o r - mance of human e x p e r t s in specific applications.A s was said above, o u r system might b e useful t o a r e g u l a t o r y agency o r regional planning a u t h o r i t y , especially in view of t h e i r typical r e s o u r c e c o n s t r a i n t s and t h e problem of maintaining technical competence vis-a-vis industry. Once developed, t h e software constituting t h e e x p e r t system c a n b e multiplied and distributed a t virtually no c o s t , and t h e hardware requirements t o s u p p o r t t h i s software are continuing t o d e c r e a s e (e.g., Fedra and Loucks, 1985).
5.2 Availability of Information
In building a computer-based information system, w e clearly imply t h a t informed decisions are "better" decisions. E x p e r t system development poses a fundamental question concerning t h e n a t u r e of knowledge, both in terms of i t s formal r e p r e s e n t a t i o n and as a n essentially social phenomenon:
knowledge as something t h a t i s s h a r e d and t r a n s f e r r e d among people and machines.
The problem, obviously, is not only one of information as such, but a l s o of interaction, communication, and of c o u r s e institutional s t r u c t u r e s . What w e are proposing is t h e development and t h e t r a n s f e r of tools and skills r a t h e r t h a n "solutions". W e want
to
build t h e modeling a p p r o a c h into t h e decision-making p r o c e s s and i t s institutional framework. This will r e q u i r e close attention t o customized design, on-site implementation, on-the-job training, and continuing s u p p o r t and maintenance.Availability of information certainly implies a c e r t a i n style, format, and e a s e of generating t h e required information quickly:
t h e u s e r may be impatient with time-consuming and s t r e s s f u l problem specification and assessment procedures;
he may want t o see a t least tentative r e s u l t s promptly, In a matter of minutes, in p a r t i c u l a r if t h e task i s sufficiently urgent;
he may lack i n t e r e s t in much of t h e underlying technical details, and not want t o directly i n t e r a c t with t h e complex quantitative procedures f o r analysis and decision support t h a t are not tailored t o t h e task s t r u c t u r e of t h e problem
at
hand;consequently, h e will r e q u i r e a format of interaction t h a t a d a p t s t o t h e style of decision making a p p r o p r i a t e t o a given task and i t s institu- tional framework.
The information provided must t h e r e f o r e include
at
least:a set of c r i t e r i a , objectives, and constraints, possibly including infor- mation on t h e relationship among objectives;
a set of alternatives (scenarios), including t h e description of basic assumptions;
a set of r e s u l t s f o r each alternative, describing i t s performance in terms of t h e above c r i t e r i a , objectives, and constraints.
5.3 Knowledge Acquisition: a B o t t l e n e c k f o r Development
The t r a n s f e r of knowledge from t h e e x p e r t system t o i t s u s e r s is a cen- t r a l problem in designing i t s interface. Transfer of knowledge from t h e human e x p e r t t o t h e system is a problem of knowledge engineering. Figure 5.1 outlines t h e basic s t r u c t u r e of t h e knowledge acquisition process. An operational system in a dynamic application domain will r e q u i r e frequent updating of i t s factual d a t a bases. Obvious items are new regulations, emerging products and production technologies, changes in t h e regional i n f r a s t r u c t u r e and land use, etc.
KNOVLEDGE BASE N E V m L I l I C E
~KMOVLEDGE
RULE
ACQUISITIOH
A I
F i g u r e 5.1: l'iuo modeLs of knowledge a c q u i s i t i o n ( M e r D a v i s a n d L e n a t , 1982).
KNOWLEDGE
4While some of these problems could be solved by connecting distributed systems into a loosely coupled network with a central site, where the updates a r e made, the region-specific information will require knowledge
)I