GIS IN ROME: FROM SOFTWARE ARCHEOLOGY TO JUBILEE_2000> PROJECT

A CASE HISTORY OF GIS ARCHEOLOGY



ACEA,  the public utility for the supply of electrical  power and

water in Rome,  launched its first project for  a graphic-digital

system to  document  its networks  in 1977  when acronyms such as

CAD  (Computer  Aided  Design),   AM/FM  (Automated  Mapping  and

Facility  Management),  and GIS (Geographic  Information Systems)

where relatively unknown.  The area served by the Municipality is

approximately  150,000  hectares   and hosts a  community  of 3,5

million.  Applicon  Graphic  System  (AGS/880)  was  the  spatial

software environment chosen,  working on  a customized DEC PDP-11

with memory management. 



For  reasons difficult to  explain -  I  would refer to  it  as a

cultural clash -,  very rarely   the local government in Italy is

able to produce,  and sell, digital cartography of its  territory

at  the   scale   necessary     for  most   utilities  (1:1,000).

Therefore,  this lack of background maps is the  puzzle that each

utility,   willing  to  build  its  digital  network  information

system,  has to  solve.  Of  course,  this fact usually makes GIS

projects extremely expensive when faced by a single company.



The first study in  ACEA  to  automate  the technical information

dates  back  to  1972.   In  1977  two  self-contained  projects,

finalized  to  the automation  of  cartographical,  graphical and

technical  data  of the  networks managed by ACEA  (1),  started:

automated mapping and microfilm technology.



Eight year later,  the   over 30,000 hectares patiently digitised

became  suddenly  inaccessible  coded data when the  AGS software

disappeared from the market.  At that  time there  were no export

routines from   a proprietary format to a standard one. Moreover,

the  PDP-11  was aging  and  the  evolution  of  the mass-storage

technology made  obsolete  the database  design  based   upon the

continuous mount  and  dismount  of  tapes  to  optimize  the  12

million character space available on the fixed disK.



In  the  meantime,  the  interactive  graphic  system  technology

evolved into  AM/FM software.  ACEA's top-management  was of  the

opinion that  investing   in  automated mapping were  useless and

decided to   remove the project manager.  No  migration  path was

planned.



The mechanical stress that the 11,000  A0-format maps documenting

ACEA's networks continuously  suffer remains an open  problem and

causes a progressive loss of information.   To limit the increase

of entrophy,  a continous work of re-drawing is   necessary, at a

cost comparable  to   the  cost  of  digitation.  In 1990 the EDP

proposed a rescue  plan:  a  migration  from  the  data  coded in

AGS/880   format to Bravo III, and from Bravo III to GDS (Graphic

Design  System).  The  idea  was  to  build  a  customized  AM/FM

environment based on GDS.



Unfortunately GDS was a  product relatively unknown  in Italy and

the closest experts were based in Holland.  The blossoming of the

GIS   technology   overshadowed   the   fortune   of   the  AM-FM

applications. The project was abandoned.



As for the future, the EDP management decided to invest only in a

GIS with the characteristic of steadiness: industry standard with

expected  life-span   past  the   year  2000,   completed   by  a

off-the-shelf  packaged software for network management.



The title of this paper -  GIS in Rome - would seem to reduce all

the   GIS applications in Rome to the experience of ACEA. This is

not completely true.  An example of successful GIS is  the raster

base for urban planning   used by  the Municipality  of Rome.  In

any case,  at the end of the Seventies and in  the  mid Eighties,

ACEA's  project was the most  ambitious  project  of  large scale

cartography in vector format made in Italy by a local utility. In

the early Nineties, the GIS of the City of Turin, a success story

of local government producing large scale background  maps of its

territory, covers only 15,000 hectares.





THE INTERPRETATIVE MODEL



The best  known model of  computing evolution in  organization is

the stage model developed by R.  Nolan between 1969 and 1979 (2).

It is based on the identification  of state  changes indicated by

changes in the budget for information technology.  Its six stages

-    Initiation,    Contagion,    Control,    Integration,   Data

Administration  and Maturity  -  shown  in  Figure  1, follow the

classical S-shaped learning curve.  In the above  case history it

is possible to highlight three main stages:



i)   Pioneering

ii)  Technocracy

iii) Cost-Effectiveness or Maturity.



Pioneering  comprises  the  Nolan's   Initiation   and  Contagion

stages:  the enthusiasm  and the lack  of knowledge   about a new

technology  very  often  cause  misjudegment  of   the  resources

necessary for a GIS and lead to time and cost overrun.



During Technocracy, the choice is made in relation to a  software

environment checked against a detailed list  of requirements. The

stress is on Control,  Integration and  Data Management; in other

words,  efficiency.  Other facets  that play important  roles are

often  underestimated:  product's  market  share,  local support,

future investment in the product.



In the Cost-Effectiveness  or Maturity stage,  strategic concerns

are  taken into  account and  the  decision  is  made  on  a more

comprehensive base.  



Unfortunately,  the learning  curve is  one-man's  experience and

cannot be taught even within  the same company.  While the ACEA's

EDP management  was dealing with  the  problems  of  the maturity

stage,  an unsatisfied user was on  his way to  re-experience the

well known errors of the Pioneers'  era: each of us has to follow

his/her karma.  A year later,  his SOS message   was the seed for

the present GIS for the water supply network.





THE PRESENT: A GIS FOR THE WATER SUPPLY NETWORK



In 1992  Nolan generalized his original proposal in a  "stage and

era"  model  that foresees three   eras:  DP (or classical  Nolan

model),  Micro,  and Network; each  era is depicted by a learning

curve and  the  DP  era  is  subdivided  into  the  six classical

stages. 



A  pilot-project was  launched  to  capitalize  on  the expertise

necessary for a  large  scale  project.  The   project, developed

between 1994  and 1996  is a low  profile, Geographic Information

System  for the   management and maintenance of the  Water Supply

Network,  integrated with a help-desk application (Water Claims).

The project's main characteristics are the following:



i)   a cartographic  base (scale 1:5,000)  surveyed and digitized

     in  the  Eighties  and  updated  between  January  1995  and

     December 1996;

ii)  GIS software ArcInfo and network management  software TecNet

     Rel.2.2 taylored to ACEA's requirements;

iii) a new help-desk (Water Claims),  a client/server application

     to manage over 20,000 claims per year.



1:5,000 scale cartography, practically useless for the management

and maintenance of  the electrical  network,  is the correct tool

for satisfying  approximately 80%  of  the  water  supply network

needs of  a large urban  area  and  was  therefore  the candidate

network for the pilot project.



The feasibility study  included  the  analysis  of  the  task and

elementary  operations,  shown  in  Figure 2,  with the empirical

measurement   of  the time  necessary  to  update  a  map  in the

traditional   paper-based   environment   and   in   the  digital

interactive graphic system.



The database was  designed by a mixed team  of  ACEA's and Esri's

analysts and,  due to the limited  availability of  key-personnel

(hydraulic engineers)  for interviews, lasted from March 1995  to

October 1995.  It was followed by a transfer of know-how in order

to enable  GIS programmers to  write the software  necessary  for

translating  the digital  network  data  from  ArcInfo  to TecNet

format.



In the pipeline there is the extension  of the GIS  to power  and

public lighting networks.  For this project,  the experimentation

of the Spatial Database Engine (SDE) technology is planned.



The Water Claims  application is  a help  desk  environment that,

using  the  workflow  approach,   follows  the  customer's  claim

navigating through the  commercial and technical functions in the

Company's organizational  chart.  The application follows a claim

from  the beginning,  usually a telephone call  to  the help desk

where skilled personnel fill out a trouble report  and follow the

claim  until the failure  is  repaired.  The  technology  is C/S:

Oracle as data base  server, and Visual Basic on the client side.





THE BENEFIT-COST MODEL



The confidence  gained  with  the pilot project is  stimulating a

more ambitious one.  ACEA is a multiservice utility and therefore

its motivation towards investing in high-quality  background maps

is higher when compared to mono-product utilities.



A  Benefit/Cost  model,  shown  in  Figure  3,  was  developed to

evaluate the effectiveness  of GIS applications.  The model, used

in  the feasibility study before  the GO-NOGO  decision, compares

the cost  of  the present paper-based system  to the  cost of the

GIS-based  system  in  the  medium  term.  The  concept  of value

broadens the concept of  benefit.  Value is  based on  the effect

information technology investment has on the business performance

of an enterprise. The model requires two obvious assumptions:



i)   the value of  the information contained in  the maps  (V) is

     greater  than the  operational cost necessary  to manage the

     present system;



ii)  the  value of  the  digital  network  information  system is

     greater  than  the  value  of  the  traditional  paper-based

     system.



In Figure 3,  once defined all the variables used  in  the model,

statement  (ii),   translated  into  the   inequality    (1),  is

algebraically  manipulated:



             V (new) > K * V (old)      (K > 1)   (1)



where the value  of the coefficient  K (K=1.5)  is the outcome of

the  requirement analysis.  The undelying hypothesis is  that the

multiservice utility uses a paper-based system. The result is the

translation into algebraic symbols of a  self-evident  truth: the

more the networks served, the more convenient for a multi service

utility  to  develop a digital  network  information  system. The

inequality  can be  solved  in   terms  of the  variable  H*, the

(unknown)  area over  which the gathering  of  data,  even  for a

single multiservice utility, is justified.





CONCLUSION



In the year 2000  Rome will host the R.  C.  Ch. Jubilee. Over 63

works-in-progresss  are planned with  a budget of 2-3  billion US

dollars.  Background  maps are   necessary for simulation studies

(traffic) and for the coordination of the different yards.



The idea  is  to  set up  a GIS factory  based  upon  large scale

background  maps  as  a  backbone  infrastructure  (GIS highway).

This  infrastructure  will  be  the  base  for  several customer-

oriented applications.



ACEA,  with its renewed experience in  GIS,   is candidated to be

the   technical Agency of the local government for the production

of the background maps and to the resolution of the  Gordian knot

of  continuously  updating the maps.  On the other  hand, ACEA is

willing to exploit the GIS technology for its core  business, not

only for its Network Information  System but also as  a dashboard

for geo-marketing and customer care.





REFERENCES

(1)  G.   Trozzi,  G.  Rondinini,  E.  Fralleoni,  Automation  of

     cartographical,  graphical and technical data of the utility

     networks of ACEA, (internal paper) 1977

(2)  R.L.  Nolan, Managing the crisis in data processing, Harvard

     Business Review, March-April 1979, pp. 115-126

(3)  E.  Orlandi,  GIS  Economics:  A Benefit  Cost  Ratio  for a

     Multi-Service Public Utility,  GIS for Business, Madrid, Feb

     1995, pp. 211-214

(4)  M.M.  Parker,  R.J.  Benson, Information Economics, Prentice

     Hall, 1988