Expert Systems for Environmental Applications - American Chemical

Chapter 12

The Cost of Remedial Action Model

Expert Systems for Environmental Applications Downloaded from pubs.acs.org by EAST CAROLINA UNIV on 01/04/18. For personal use only.

Expert System Applications Marie T. Chenu and Jacqueline A. Crenca CH2M Hill, P.O. Box 4400, Reston, VA 22090 The Cost of Remedial Action (CORA) model was developed for the United States Environmental Protection Agency (EPA). The model i s used i n developing and costing remedial actions for Superfund sites before or during the remedial investigation of the cleanup. The CORA model includes two independent, microcomputer-based subsystems. One subsystem i s a knowledge-based consultation program that develops remediation recommendations. The second subsystem i s a database management system that develops site-specific cost estimates for the technologies required to implement the expert system's recommendations. Use of the model has made possible a considerable time saving over manual scoping and costing, and has also led to consistent procedures for remedy selection across EPA regions. This paper discusses the expert system anditsdevelopment, testing, validation, and application. The Cost o f Remedial A c t i o n (CORA) model was d e v e l o p e d f o r t h e U n i t e d S t a t e s E n v i r o n m e n t a l P r o t e c t i o n Agency (EPA). The model i s used i n d e v e l o p i n g and c o s t i n g r e m e d i a l a c t i o n s f o r Superfund s i t e s before o r during the remedial i n v e s t i g a t i o n of the cleanup. System

Capabilities

The CORA model was d e s i g n e d t o r u n on an IBM-compatible microcomp u t e r . I t r e q u i r e s 587K o f f r e e RAM (beyond t h e RAM r e q u i r e d by t h e o p e r a t i n g system) and a minimum o f 5 megabytes o f f r e e d i s k s p a c e . 0097-6156/90/0431-0162$06.00/0 © 1990 American Chemical Society


12.

CHENU AND CRENCA

The Cost ofRemedial Action Model

163

The model runs i n e i t h e r a c o l o r o r a monochrome mode, depending on the m o n i t o r . The CORA model i n c l u d e s two i n d e p e n d e n t , microcomputer-based subsystems. One subsystem i s a knowledge-based c o n s u l t a t i o n program d e v e l o p e d w i t h t h e L e v e l 5 e x p e r t system s h e l l , dBASE I I I + , and Nantucket C l i p p e r . T h i s subsystem c o m p r i s e s f o u r knowledge bases t h a t communicate w i t h each o t h e r and update f a c t s d u r i n g e x e c u t i o n . The second subsystem i s a database management system w r i t t e n i n dBase I I I + and Nantucket C l i p p e r t h a t d e v e l o p s s i t e - s p e c i f i c c o s t e s t i m a t e s f o r t h e t e c h n o l o g i e s r e q u i r e d t o implement t h e e x p e r t system's recommendations. The e s t i m a t e i s t h a t use o f t h e model has made p o s s i b l e a 5 - f o l d t o 1 5 - f o l d time s a v i n g o v e r manual s c o p i n g and c o s t i n g . Use of t h e model has a l s o l e d t o c o n s i s t e n t p r o c e d u r e s f o r remedy s e l e c t i o n a c r o s s EPA r e g i o n s . The e x p e r t system a n a l y z e s a s i t e by f o c u s i n g on s e p a r a t e u s e r d e f i n e d c o n t a m i n a t e d a r e a s of t h e s i t e . F o r each c o n t a m i n a t e d a r e a , the u s e r may d e s i g n a t e up t o 13 waste t y p e s , from b u r i e d drums t o contaminated saturated s o i l s . A f t e r an i n t e r a c t i v e d i a l o g w i t h t h e u s e r , t h e e x p e r t system produces a summary r e p o r t t h a t l i s t s a range of p o t e n t i a l l y implementable and a p p l i c a b l e r e m e d i a l a c t i o n (RA) t e c h n o l o g i e s f o r each waste t y p e . The t e c h n o l o g i e s , w h i c h i n c l u d e b o t h t r e a t m e n t and containment t e c h n o l o g i e s and range from u s i n g vapor-phase c a r b o n t o a s p h a l t c a p s , c a n be combined by t h e u s e r t o form one o r more RA a l t e r n a t i v e s . F o r each q u e s t i o n asked by t h e e x p e r t system, t h e u s e r i s p r e sented w i t h a menu, o r l i s t , of p o s s i b l e answers ( m a i n l y i n "True/ F a l s e " f o r m a t ) from which t o choose, o r t h e u s e r may be asked t o e n t e r t e x t u a l i n f o r m a t i o n , u s i n g t h e k e y b o a r d . The u s e r i n t e r f a c e also includes extensive on-line help f o r f i r s t - t i m e or infrequent u s e r s who may n o t be f a m i l i a r w i t h e x p e r t system t e r m i n o l o g y o r w i t h the u s e r i n t e r f a c e of t h e L e v e l 5 e x p e r t system s h e l l . T h i s " h e l p " f e a t u r e c a n be i n v o k e d a t any p o i n t d u r i n g a c o n s u l t a t i o n by p r e s s ing a f u n c t i o n key. A f t e r RA s c e n a r i o s a r e d e t e r m i n e d f o r t h e s i t e , t h e c o s t system d e v e l o p s e s t i m a t e s t h a t have a t a r g e t a c c u r a c y range o f +50 t o -30 percent o f a c t u a l c o s t s . The e x p e r t system c o m p r i s e s 40 t e c h nologies covering containment, removal, treatment, and d i s p o s a l o p t i o n s . I n a d d i t i o n t o c o n t a i n i n g c o s t a l g o r i t h m s f o r each o f t h e t e c h n o l o g i e s t h a t may be recommended by t h e e x p e r t system, t h e c o s t system d e v e l o p s c o s t e s t i m a t e s f o r s i t e p r e p a r a t i o n , s i t e a d m i n i s t r a t i o n , h e a l t h and s a f e t y , and c o n t i n g e n c i e s and a l l o w a n c e s . The system i s not d e s i g n e d t o i n c o r p o r a t e a l l of t h e many t e c h n o l o g i e s t h a t would be n e c e s s a r y t o a d d r e s s e v e r y type o f s i t e ; i n s t e a d , the goal i s t o address the m a j o r i t y of s i t e s . "Outliers" i n c l u d e m i n i n g s i t e s and s i t e s c o n t a i n i n g r a d i o a c t i v e waste. Some emerging t e c h n o l o g i e s , such as i n s i t u v i t r i f i c a t i o n o r u l t r a v i o l e t

164

E X P E R T SYSTEMS FOR ENVIRONMENTAL APPLICATIONS

o z o n a t i o n , were n o t i n c l u d e d i n t h e model because o f t h e i r scope and u n c e r t a i n t i e s . The CORA framework, however, a l l o w s f o r e x p a n s i o n s , and o t h e r t e c h n o l o g i e s w i l l be c o n s i d e r e d f o r a d d i t i o n d u r i n g a n n u a l updates o f t h e model.


Development

o f t h e E x p e r t System

Overview. The CORA e x p e r t system has gone t h r o u g h two major phases i n i t s development. The f i r s t - p h a s e d e l i v e r a b l e was a p r o t o t y p e system. The knowledge bases and e x p l a n a t i o n f a c i l i t y were w r i t t e n u s i n g t h e P r o d u c t i o n R u l e Language (PRL) o f I n s i g h t 2+. A d d i t i o n a l user i n t e r f a c e and r e p o r t i n g capabilities were d e v e l o p e d i n dBASE I I I + and Nantucket C l i p p e r . The p r o t o t y p e c o n t a i n e d 550 r u l e s i n s i x knowledge bases d e a l i n g w i t h c o n t a m i n a t i o n i n groundwater, lagoons and ponds, and soil. The p r o t o t y p e was d e s i g n e d t o be used by a CH2M HILL r e p r e s e n t a t i v e t r a i n e d i n u s i n g t h e model and knowledgeable about Superfund s i t e s , w o r k i n g w i t h an EPA r e g i o n a l p r o j e c t manager (RPM). On t h e b a s i s o f an i n t e r a c t i v e c o n s u l t a t i o n , t h e system c o u l d recommend up t o 28 r e m e d i a t i o n t e c h n o l o g i e s . I n t h e second phase o f t h e p r o j e c t , EPA wanted t o d i s t r i b u t e the model t o t h e EPA r e g i o n s f o r independent u s e , and CORA was r e f i n e d and upgraded t o f a c i l i t a t e t h a t use. The e x p e r t system was r e v i s e d t o i n c l u d e f o u r knowledge bases c o n t a i n i n g 671 r u l e s and up t o 40 t e c h n o l o g i e s f o r recommendation. I n s i g h t 2+, whose name was changed t o L e v e l 5; dBASE I I I + ; and t h e Nantucket C l i p p e r c o m p i l e r were used. The second v e r s i o n o f t h e e x p e r t system was completed i n a p p r o x i m a t e l y 8 months and was r e l e a s e d i n June 1988. EPA and t h e U.S. Navy have used i t t o s e l e c t and c o s t RAs under t h e i r f i s c a l y e a r 1989, 1990, and 1991 budgets f o r r e m e d i a t i n g hazardous waste sites. A t h i r d v e r s i o n o f CORA i s s c h e d u l e d f o r r e l e a s e i n e a r l y 1990. I t w i l l c o n t a i n updates o f b o t h t h e e x p e r t system and t h e c o s t system. System G o a l s . From t h e b e g i n n i n g , t h i s p r o j e c t was shaped by s e v e r a l important c o n s t r a i n t s . F i r s t , t h e system's recommendations had t o be based on c u r r e n t EPA p o l i c i e s and t e c h n o l o g i c a l c o n s i d e r a t i o n s , so t h e system had t o be d e s i g n e d t o a l l o w changes as b o t h p o l i c i e s and t e c h n o l o g i e s e v o l v e . This constraint s i g n i f i c a n t l y i n f l u e n c e d o u r d e c i s i o n t o s e l e c t a k n o w l e d g e - r e p r e s e n t a t i o n scheme based on p r o d u c t i o n r u l e s , where each r u l e i s c o n s t r u c t e d i n a simple " I f . . . then . . . " f o r m a t , t h e premise b e i n g a B o o l e a n e x p r e s s i o n and t h e a c t i o n c o n t a i n i n g one o r more c o n c l u s i o n s . Each r u l e i s modular and independent o f t h e o t h e r s . From a development


12.

C H E N U AND C R E N C A


165

s t a n d p o i n t , such modular c o d i n g l e a d s t o a r e l a t i v e l y u n c o m p l i c a t e d c o n t r o l s t r u c t u r e (1) t h a t has the b e n e f i t of a l l o w i n g new r u l e s t o be added e a s i l y . The second c o n s t r a i n t was t h a t the f i r s t p r o t o t y p e needed t o be ready f o r f i e l d t e s t i n g i n l e s s t h a n 6 months. T h i s c o n s t r a i n t l e d us t o choose an e x p e r t system s h e l l , because a s h e l l would have the i n f e r e n c e engine and u s e r i n t e r f a c e b u i l t i n , p e r m i t t i n g s i g n i f i c a n t time s a v i n g i n comparison t o the time r e q u i r e d f o r i m p l e m e n t i n g an e x p e r t system w i t h a language such as L i s p or P r o l o g . Our t h i r d c o n s t r a i n t was the need t o d e s i g n a system t h a t would run on a microcomputer w i t h o n l y the b a s i c 640K of RAM. T h i s need stems from two f a c t s : Most of EPA*s microcomputers had t h a t h a r d ware c o n f i g u r a t i o n , and EPA wanted the maximum number of u s e r s t o have a c c e s s t o CORA w i t h o u t i n c u r r i n g the h i g h c o s t of o p t i o n - l a d e n microcomputer hardware or the c o n n e c t - t i m e c o s t s and p o s s i b l e l a c k of a c c e s s of a mainframe system. F i n a l l y , the f o u r t h c o n s t r a i n t was t h a t the chosen s o f t w a r e must not l e v y l i c e n s i n g f e e s on p r o d u c t i o n copies. To meet t h e s e c o n s t r a i n t s , we chose f o r our development t o o l I n s i g h t 2+, a r u l e - b a s e d and microcomputer-based e x p e r t system s h e l l t h a t met most of the c r i t e r i a of a good t o o l , as d e s c r i b e d by Waterman ( 2 ) — t h a t i s , i t has good u s e r i n t e r f a c e , r u l e - t r a c i n g , and debugging c a p a b i l i t i e s . I n a d d i t i o n , I n s i g h t 2+ has no r e q u i r e m e n t s for l i c e n s i n g fees. Knowledge A c q u i s i t i o n and I m p l e m e n t a t i o n . The domain knowledge i n CORA was o b t a i n e d from the e x p e r i e n t i a l knowledge of CH2M HILL*s e n v i r o n m e n t a l e n g i n e e r i n g e x p e r t s , i n f o r m a t i o n on c u r r e n t EPA p o l i c i e s d e r i v e d from i n t e r v i e w s w i t h s e n i o r EPA managers, and from EPA r e p o r t s ( 3 ) , the Superfund Amendments and Reauthorization Act (SARA), and the Hazardous and S o l i d Waste Amendments (HSWA). The approach of a c q u i r i n g and i m p l e m e n t i n g knowledge t h a t was used i n the f i r s t phase of development was c u r t a i l e d by time c o n straints. I n d e v e l o p i n g the n e c e s s a r y r u l e s , knowledge e n g i n e e r s r e l i e d p r i m a r i l y on d e c i s i o n t r e e s drawn a f t e r i n t e r v i e w i n g the domain e x p e r t s . Most of the r u l e s were l i n k e d i n some way. This approach r e s u l t e d i n a v e r y f a s t system t h a t f u n c t i o n e d much l i k e a c o n v e n t i o n a l program and e x e r t e d s t r o n g c o n t r o l o v e r the i n f e r e n c e e n g i n e . P r e d i c t a b l y , the system was apt t o f a i l whenever the u s e r t r i e d t o b r a n c h t o a p a t h not d e f i n e d by the r u l e s . I n the second phase, a l e s s d e t e r m i n i s t i c approach t o knowledge i m p l e m e n t a t i o n was chosen. A s e t of p r e l i m i n a r y r u l e s i n the form of " I f . . . t h e n . . . " s t a t e m e n t s was d e v e l o p e d , and the s e t was grouped i n t o c a t e g o r i e s such as l a n d f i l l s , above-ground c o n t a m i n a t i o n , and removal o p t i o n s and i s s u e s . The knowledge e n g i n e e r s rewrote the r u l e s u s i n g PRL s y n t a x , and t h e s e new r u l e s formed the


166

EXPERT SYSTEMS FOR ENVIRONMENTAL APPLICATIONS

knowledge b a s e s . Working c l o s e l y w i t h t h e domain e x p e r t s , t h e knowledge e n g i n e e r s t h e n r e f i n e d t h e e x i s t i n g r u l e s and added new r u l e s by e x t e n s i v e l y t e s t i n g and r u n n i n g t h e knowledge b a s e s . The advantages of u s i n g a r u l e - b a s e d system s i m p l i f i e d t h e communication between knowledge e n g i n e e r s and domain e x p e r t s . Because r u l e s a r e a s t r a i g h t f o r w a r d way o f e x p r e s s i n g knowledge, u n d e r s t a n d i n g them d i d not r e q u i r e knowledge of a programming language. By f o l l o w i n g t h e i n f e r e n c e c h a i n and l o o k i n g a t each r u l e , t h e e x p e r t s c o u l d e a s i l y u n d e r s t a n d t h e system's l i n e o f r e a s o n i n g and c o u l d suggest m o d i f i cations during t e s t i n g . A major d i f f i c u l t y e n c o u n t e r e d d u r i n g i m p l e m e n t a t i o n was t h e way i n w h i c h recommendations had t o be p r e s e n t e d t o t h e u s e r s . F o r the system t o be as u s e r f r i e n d l y as p o s s i b l e , system recommenda t i o n s had t o be grouped i n t o c a t e g o r i e s such as t r e a t m e n t , c o n t a i n ment, and d i s c h a r g e , r a t h e r t h a n i n t h e o r d e r i n w h i c h t h e recommendations were deduced. A n o t h e r r e p o r t i n g c r i t e r i o n was t h a t m u l t i p l e recommendations f o r RA t e c h n o l o g y t h a t f a l l under one c a t e g o r y h e a d i n g be shown i n t h e " E i t h e r . . . o r . . ." f o r m a t . Shown below i s an a c t u a l e x p e r t system recommendation under t h e h e a d i n g "In S i t u S o i l Treatments": Either ο S o i l - v a p o r e x t r a c t i o n f o r VOCs Either ο F l a r e f o r VOCs Or ο Vapor-phase c a r b o n Or ο Deal w i t h t h e p r i n c i p a l t h r e a t ( i . e . , b u r i e d t a n k s o r drums) L e v e l 5 had no f a c i l i t y t o h a n d l e t h e s e r e p o r t i n g r e q u i r e m e n t s a d e q u a t e l y , so a s i g n i f i c a n t p a r t o f t h e i m p l e m e n t a t i o n p e r i o d was spent w o r k i n g on t h e s t r a t e g y f o r r e p r e s e n t i n g such " n e s t e d " l e v e l s of recommendations. A d d i t i o n a l r u l e s had t o be added t o w r i t e t h e recommendations t o d i s k , and dBASE 111+ programs were w r i t t e n t o s o r t and p r e s e n t t h e recommendations t o the u s e r s . See F i g u r e 1 f o r an o v e r v i e w o f t h e e x p e r t system a r c h i t e c t u r e . The CORA e x p e r t system was d e v e l o p e d by a team o f t h r e e knowledge e n g i n e e r s , two o f whom were programmers and one o f whom was a domain e x p e r t . The l a t t e r was r e s p o n s i b l e f o r i n t e r v i e w i n g o t h e r domain e x p e r t s . Numerous o t h e r domain e x p e r t s a l s o s e r v e d as r e v i e w e r s . An example of o u t p u t from an e x p e r t system r u n i s shown i n F i g u r e 2.


C H E N U AND CRENCA

F i g u r e 1.


Overview of CORA E x p e r t

System



CORA EXPERT SYSTEM

RUN: TEST RUN FOR ACS RUN BY: M. CHENU SITE: TEST SITE CONTAMINATED AREA: CONTAMINATED AREA 1 WASTE TYPE: HOT SPOTS (UNSATURATED MTL AROUND LEAKY TANKS OR DRUMS) INPUT Response t y p e : Treatments S o i l d e s c r i p t i o n : Medium sand S o i l s around drums o r t a n k s : True S o i l c o n t a m i n a n t : V o l a t i l e o r g a n i c compounds VOCs i n s o i l s pose t h e p r i m a r y r i s k : False Excavation acceptable: True Site conditions could threaten: False Exposed t o e r o s i o n : F a l s e RECOMMENDATIONS FOR HOT SPOTS (UNSATURATED MTL AROUND LEAKY TANKS OR DRUMS) GENERAL ο 503 Groundwater

monitoring

IN SITU SOILS TREATMENTS Either ο 305 S o i l vapor e x t r a c t i o n f o r VOCs Either ο 306 F l a r e f o r VOCs Or ο 308 Vapor phase c a r b o n Or ο Deal w i t h the p r i n c i p a l t h r e a t ( i . e . b u r i e d t a n k s o r drums) REMOVAL OPTIONS ο 201 S o i l e x c a v a t i o n F i g u r e 2.

Example o f CORA E x p e r t

System Output


12.



169

Knowledge Bases. As shown i n F i g u r e 3, t h e second phase CORA e x p e r t system has f o u r knowledge b a s e s . The f i r s t knowledge b a s e — C O R A — i s a s m a l l one o f 15 r u l e s t h a t s i m p l y a s k s t h e u s e r t o s p e c i f y t h e waste t y p e s f o r each c o n t a m i n a t e d a r e a , t h e n c a l l s t h e second knowledge b a s e — M A I N — t o examine each waste t y p e . MAIN has 492 r u l e s t h a t a r e grouped i n t o t h e f o l l o w i n g c a t e g o r i e s : removal, t r e a t m e n t , c o n t a i n m e n t , l a n d f i l l , above-ground c o n t a m i n a t i o n , n a t u r a l a t t e n u a t i o n , and a c t i v e r e s t o r a t i o n . The r u l e s examine a l l c o n t a m i n a n t s s p e c i f i e d by t h e u s e r and t r y t o recommend s u i t a b l e RA t e c h n o l o g i e s . The t h i r d knowledge b a s e — L A N D F I L L — c o n t a i n s 71 r u l e s and d e a l s e x c l u s i v e l y w i t h l a n d f i l l i s s u e s f o r b y - p r o d u c t s g e n e r a t e d by t r e a t m e n t o r c o n t a i n m e n t . The f o u r t h knowledge b a s e — W A T E R — h a s 43 r u l e s and d e a l s w i t h t h e t r e a t m e n t o f l i q u i d s g e n e r a t e d by t h e t r e a t m e n t o r containment recommended by MAIN o r LANDFILL. To m i n i m i z e t h e number o f r u l e s and t h u s reduce t h e RAM r e q u i r e d t o r u n t h e e x p e r t system, we t o o k f u l l advantage o f t h e c h a i n i n g c a p a b i l i t y o f L e v e l 5. C h a i n i n g c a p a b i l i t y i s t h e a b i l i t y of knowledge bases t o c a l l each o t h e r . F o r example, once t h e MAIN knowledge base d e t e r m i n e s t h a t a l a n d f i l l may be an o p t i o n f o r a waste b y - p r o d u c t , t h e LANDFILL knowledge base i s c a l l e d and t h e r e l e v a n t f a c t s a r e passed t o i t t h r o u g h a t e x t f i l e . The f a c t s a r e t h e n updated and passed back t o t h e c a l l i n g program. This e l i m i n a t e s redundant c o d i n g because each subsystem d e a l s w i t h o n l y a s m a l l subset o f t h e domain knowledge. The i n c r e a s e d e f f i c i e n c y , however, i s a c q u i r e d a t t h e c o s t o f a l o s s o f c o n t e x t knowledge by the e x p e r t system as a whole. Because each knowledge base d e a l s w i t h o n l y a s m a l l s e t of f a c t s , t h e u s e r cannot r e s t a r t t h e system at an a r b i t r a r y p o i n t ( f o r example, d u r i n g t h e e x e c u t i o n o f t h e LANDFILL knowledge b a s e ) , change a p r e v i o u s answer, and r e s t a r t t h e system from t h a t p o i n t w i t h o u t l o s i n g t h e c o n t e x t o f t h e c o n s u l t a tion session. Another problem we e n c o u n t e r e d w i t h t h e L e v e l 5 s o f t w a r e was i t s e r r a t i c b e h a v i o r when a s m a l l knowledge base such as CORA c h a i n s t o a l a r g e knowledge base such as MAIN. G l o b a l f a c t s t h a t s h o u l d be s h a r e d were n o t passed t o MAIN. To work around t h i s problem, we had t o w r i t e t h e s e f a c t s t o a t e x t f i l e t h a t i s read by MAIN a t s t a r t up. R u l e s and C o n f i d e n c e F a c t o r s . F a c t s t h a t a r e known t o t h e system are g i v e n a c o n f i d e n c e l e v e l o f e i t h e r 0 ( n o t t r u e ) o r 100 ( t r u e ) . F a c t s t h a t a r e n o t y e t known t o t h e system a r e a s s i g n e d a c o n f i d e n c e l e v e l o f - 1 , and f a c t s t h a t a r e d e c l a r e d unknown a r e g i v e n a c o n f i d e n c e l e v e l o f - 2 . CORA was n o t s t r u c t u r e d t o a s k u s e r s f o r a degree o f c o n f i d e n c e i n t h e i r answers. Each r u l e ' s c o n c l u s i o n i s a s s i g n e d e i t h e r 0 i f t h e r u l e f a i l s o r 100 i f t h e r u l e s u c c e e d s . I n t h a t r e s p e c t , CORA i s not d e s i g n e d t o d e a l w i t h " f u z z y r e a s o n i n g . "


E X P E R T SYSTEMS F O R ENVIRONMENTAL APPLICATIONS

MAIN

General, removal, and initial treatment and discharge issues are addressed

If done with waste type

If landfill is recommended LANDFILL

Deal with landfill options and issues

Display recommendations on screen and generate a hard copy of the user input and recommendations If water treatment required WATER

If onsite Deal with water treatment options and issues landfill required

F i g u r e 3. E x p e r t System:

Program Flow


12.

C H E N U AND CRENCA


171

There a r e , however, r u l e s t h a t check f o r unknown f a c t s and t r y t o pursue a f a i l e d l i n e of r e a s o n i n g i f the l i n e of r e a s o n i n g i s n e c e s s a r y f o r a c h i e v i n g the c u r r e n t g o a l . To i n c r e a s e the r o b u s t n e s s of the system, the knowledge e n g i n e e r s have implemented two ways of d e a l i n g w i t h u n c e r t a i n t y reasoning. One method i s t o use m u l t i p l e r u l e s w i t h the same conclusion. For example, one of the g o a l s when d e a l i n g w i t h groundwater i s t o determine the v a l u e of the h y d r a u l i c c o n d u c t i v i t y of the s a t u r a t e d zone. I f the u s e r i n d i c a t e s i n response t o a system q u e s t i o n t h a t t h i s v a l u e i s unknown, o t h e r r u l e s w i t h the same g o a l w i l l ask the u s e r t o choose from a l i s t of 12 a t t r i b u t e s the one t h a t d e s c r i b e s the s a t u r a t e d zone i n q u e s t i o n (such as c l a y , g r a v e l , s a n d ) . On the b a s i s of t h i s s e l e c t i o n , the system a s s i g n s a d e f a u l t h y d r a u l i c c o n d u c t i v i t y v a l u e t o the zone and proceeds t o the next r u l e . Another method of d e a l i n g w i t h u n c e r t a i n t y r e a s o n i n g i s t o check e x p l i c i t l y f o r f a c t s t h a t have a c o n f i d e n c e l e v e l of -2. For example, i f the contaminated a r e a i s groundwater, the u s e r w i l l be asked t o s e l e c t e i t h e r n a t u r a l a t t e n u a t i o n or a c t i v e r e s t o r a t i o n as a response a c t i o n . R u l e s f o r c h e c k i n g the c o n f i d e n c e l e v e l of the response a c t i o n are i n p l a c e i n the knowledge b a s e s . I f the d e t e r m i n a t i o n i s t h a t the f a c t has a v a l u e of -2, o t h e r r u l e s are i n p l a c e f o r d e t e r m i n i n g t h i s g o a l t h r o u g h a s e r i e s of a l t e r n a t i v e questions. Control Structure. The r u l e s are i n v o k e d i n a b a c k w a r d - c h a i n i n g scheme t h a t produces a d e p t h - f i r s t s e a r c h — t h a t i s , the i n f e r e n c e engine t r i e s t o s a t i s f y a g o a l by e v a l u a t i n g a l l r u l e s t h a t l e a d t o t h a t g o a l . H e u r i s t i c r u l e s were used as much as p o s s i b l e t o s h o r t e n the s e a r c h . For example, the system may ask the u s e r t o s e l e c t the d e s i r e d response a c t i o n , e i t h e r containment or t r e a t m e n t of the wastes. I f the user chooses c o n t a i n m e n t , r u l e s f o r t r e a t i n g the main wastes w i l l not be a c t i v a t e d . I f b y - p r o d u c t s of the c o n t a i n ment t e c h n o l o g i e s r e q u i r e t r e a t m e n t , however, r u l e s f o r t r e a t m e n t w i l l s t i l l be e x p l o r e d . I n t e r f a c e w i t h E x t e r n a l Programs. One drawback of L e v e l 5 i s t h a t the t r a c i n g c a p a b i l i t i e s e x i s t o n l y i n the p r o d u c t i o n environment. Compiled v e r s i o n s of the knowledge bases do not a l l o w u s e r s t o see the l i n e of r e a s o n i n g or t o change or look at f a c t s p r o c e s s e d d u r i n g a consultation. So t h a t u s e r s can keep a r e c o r d of t h e i r c o n s u l t a t i o n , a d d i t i o n a l u s e r - i n t e r f a c e and r e p o r t i n g c a p a b i l i t i e s were added v i a dBASE I I I + programs t h a t are a c t i v a t e d from w i t h i n the knowledge b a s e s . When the system r e c e i v e s a u s e r response d u r i n g a s e s s i o n , the system q u e s t i o n and u s e r answer are w r i t t e n t o d i s k . When the

172



system a r r i v e s a t a recommendation, t h e recommendation i s a l s o written to disk. Such i n t e r n a l bookkeeping added g r e a t l y t o t h e number o f r u l e s and t h e c o m p l e x i t y of t h e knowledge b a s e s . At t h e end o f t h e s e s s i o n , t h e dBASE I I I + program i s l o a d e d i n t o memory t o g a t h e r and s o r t t h e d a t a f i l e . The program d i s p l a y s the system's recommendations on t h e s c r e e n and g i v e s t h e u s e r t h e o p t i o n of having a p r i n t e d record of the s e s s i o n . This printed r e c o r d would i n c l u d e t h e system's q u e s t i o n s , t h e u s e r ' s answers, and the system's recommendations. T e s t i n g and V a l i d a t i o n . B o t h t h e f i r s t and second CORA p r o t o t y p e s were t e s t e d by in-house domain e x p e r t s b e f o r e b e i n g f i e l d t e s t e d . A f t e r each m o d i f i c a t i o n o f t h e system, t e s t c a s e s were r e r u n t o ensure t h a t t h e system f u n c t i o n e d c o r r e c t l y . I n May 1987, t h e f i r s t p r o t o t y p e was f i e l d t e s t e d on 97 U. S. EPA Superfund s i t e s l i k e l y t o be FY 1989 RA c a n d i d a t e s . F o r each s i t e , CH2M HILL team members worked one-on-one w i t h U. S. EPA RPMs and completed runs o f t h e CORA e x p e r t system and t h e c o s t system. The second v e r s i o n o f CORA was a l s o t e s t e d by a s e l e c t e d group o f EPA RPMs and CH2M HILL e x p e r t s b e f o r e i t was r e l e a s e d . As a r e s u l t o f t h e t e s t i n g , a d d i t i o n a l e x p l a n a t o r y i n f o r m a t i o n was added t o t h e r u l e s t h a t might be confusing to users. The system was a l s o m o d i f i e d t o i n f o r m t h e u s e r s o f i n t e r m e d i a t e recommendations. Our two approaches t o i m p l e m e n t a t i o n p r e s e n t e d an i n t e r e s t i n g c o n t r a s t i n r u l e m o d i f i c a t i o n and debugging d u r i n g t h e t e s t i n g c y c l e . We found t h a t t h e approach i n t h e f i r s t p h a s e — w r i t i n g r u l e s t h a t e x p l i c i t l y c a l l e d each o t h e r — m a d e debugging easy but r u l e modification d i f f i c u l t . Because each r u l e was l i n k e d t o a n o t h e r , i n s e r t i n g o r d e l e t i n g r u l e s meant t h a t a l l p o s s i b l e r u l e l i n k s had t o be examined and extended o r t r u n c a t e d . The i n v e r s e was t r u e i n the second phase, where o u r approach o f w r i t i n g modular r u l e s made r u l e i n s e r t i o n easy and debugging more c o m p l i c a t e d because t h e o r d e r i n which r u l e s were f i r e d was n o t e a s i l y t r a c e a b l e . A f t e r the r e l e a s e o f t h e second phase CORA model, EPA r e t a i n e d an o u t s i d e c o n s u l t a n t t o conduct a v a l i d a t i o n s t u d y o f t h e model. The s t u d y , conducted i n J a n u a r y 1989, i n c l u d e d a r e v i e w o f t h e d e c i s i o n r u l e s and o f t h e e x p e r t system's o p e r a t i o n a l recommendations. The s t u d y c o n c l u d e d t h a t t h e e x p e r t system i s a " n e c e s s a r y and u s e f u l t o o l " ( I C F , I n c . Performance E v a l u a t i o n o f Cost o f Remedial A c t i o n (CORA) Mode 1, p r e p a r e d f o r t h e U.S. E n v i r o n m e n t a l P r o t e c t i o n Agency's Hazardous S i t e C o n t r o l D i v i s i o n , J a n u a r y 13, 1989) and c a p t u r e s t h e d e c i s i o n p r o c e s s used i n t h e S u p e r f u n d r e m e d i a l program. A p p l i c a t i o n s o f t h e E x p e r t System The CORA e x p e r t and c o s t systems have been used i n a v a r i e t y o f applications. The model was used t o scope and h e l p d e v e l o p b o t h EPA's and t h e Navy's r e m e d i a t i o n budgets f o r FY 1989, FY 1990, and FY 1991.


12.



173

CORA was a l s o used f o r r e g u l a t o r y support f o r t h e Resource C o n s e r v a t i o n and Recovery A c t (RCRA) and f o r a n a l y z i n g c o r r e c t i v e a c t i o n s t r a t e g i e s and c o s t s f o r t h e RCRA L o c a t i o n Standards R u l e . The model i s b e i n g used t o s c r e e n and e v a l u a t e t e c h n o l o g i e s and r e m e d i a t i o n s t r a t e g i e s f o r t h e Department o f Defense. The CORA model has a l s o been used t o s c r e e n t e c h n o l o g i e s , develop a l t e r n a t i v e s , and e s t i m a t e i n i t i a l r e m e d i a t i o n c o s t s . To d a t e , more t h a n 200 c o p i e s o f t h e model have been d i s t r i b u t e d t o f e d e r a l and s t a t e a g e n c i e s , f o r e i g n governments, e n v i r o n m e n t a l con s u l t a n t s , and i n d u s t r y . Both t h e CORA e x p e r t system and t h e c o s t system were d e s i g n e d t o a l l o w r e v i s i o n and e x p a n s i o n . EPA funds have been a p p r o p r i a t e d f o r c o n t i n u e d maintenance, enhancement, and i n c o r p o r a t i o n o f u s e r feedback t o r e f l e c t c u r r e n t r e g u l a t o r y p o l i c i e s , demonstrated t e c h n o l o g i e s , and c o s t c o n s i d e r a t i o n s . Use o f t h e CORA model i s expected t o c o n t i n u e . F u t u r e a p p l i cations include: ο

Use by EPA and o t h e r f e d e r a l a g e n c i e s i n e v a l u a t i n g r e m e d i a t i o n s t r a t e g i e s and i n d e v e l o p i n g f i s c a l o u t y e a r budgets

ο

Use by f e d e r a l a g e n c i e s , s t a t e s , i n d u s t r y , and e n v i r o n mental p r o f e s s i o n a l s i n s c r e e n i n g t e c h n o l o g i e s and e v a l u a t i n g remedial a l t e r n a t i v e s

ο

Use by s t a t e s f o r t o t a l - p r o g r a m and s i t e - s p e c i f i c reme d i a t i o n budgets and s c o p i n g

ο

Use i n a n t i c i p a t i n g c h o i c e s of r e m e d i a l t e c h n o l o g y and c o s t e f f e c t s f o r r e g u l a t o r y - i m p a c t a n a l y s e s o f new environmental r e g u l a t i o n s

Summary We b e l i e v e CORA was a s u c c e s s because o u r emphasis f o r c o s t and time was mainly on c o n c e p t u a l i z a t i o n , f o r m a l i z a t i o n , and knowledge a c q u i s i t i o n , as i n d i c a t e d by F i g u r e 4. D e s p i t e t h e d i f f i c u l t i e s encountered i n t h e L e v e l 5 s o f t w a r e , we b e l i e v e i t was an a p p r o p r i a t e t o o l f o r d e v e l o p i n g t h e e x p e r t system because i t r e q u i r e d v e r y l i t t l e RAM t o o p e r a t e i n c o m p a r i s o n t o most o t h e r s h e l l s on t h e market. The L e v e l 5 s o f t w a r e a l l o w e d us t o meet o u r g o a l o f d e v e l o p i n g a system t h a t would r u n on a 640K microcomputer. In a d d i t i o n , t h e L e v e l 5 programming language i s r e l a t i v e l y easy t o l e a r n , and t h e t e c h n i c a l support f o r i t was e x c e l l e n t .

EXPERT SYSTEMS FOR ENVIRONMENTAL APPLICATIONS


174

Conceptualization Formalization Knowledge Acquisition (0

Implementation

n CL C

φ E

Field Testing

Q. O Φ > Φ Q

42

3

4 Months

F i g u r e 4.

Phase 2 E x p e r t

System:

Developmental

Timeline

12.

CHENU AND CRENCA


175

Acknow1edgments CORA was d e v e l o p e d under EPA c o n t r a c t number 68-01-7090. The a u t h o r s would e s p e c i a l l y l i k e t o thank K i r b y B i g g s and Russ Wyer o f EPA*s Hazardous S i t e C o n t r o l D i v i s i o n f o r t h e i r h e l p and s u p p o r t d u r i n g t h e development of t h e model.


Literature Cited 1.

Brachman, Ronald J.; Levesque, Hector J. Readings i n Knowledge Representation; Morgan Kaufmann Publishers, Inc.: Los Altos, California, 1985.

2.

Waterman, D. A. Guide to Expert Systems; Addison-Wesley: Reading, Massachusetts, 1986.

3.

Guidance on Remedial Actions for Contaminated Groundwater at Superfund Sites, U.S. Environmental Protection Agency, Hazardous Site Control Division, Office of Emergency and Remedial Response, 1988. RECEIVED January 18, 1990

Expert Systems for Environmental Applications - American Chemical

Recommend Documents