USIB COMMITTEE ON DOCUMENTATION MINUTES OF THE EIGHTY-FOURTH MEETING, 14 DECEMBER 1966

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80B01139A000100130001-8
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 33
• Document Creation Date: December 16, 2016
• Document Release Date: February 23, 2005
• Sequence Number: 1
• Publication Date: December 22, 1966
• Content Type: MIN

Approved Felease 2005/03/16: CIA-RDP80B01 A00010013
CODIB-M-84
22 December 1966
U N I T E D S T A T E S I N T E L L I G E N C E B 0 A R D
COMMITTEE ON DOCUMENTATION
Minutes of the Eight- FFourth Meeting, 14 December 1966
Members or Their Representatives Present
25X1 Chairman
CIA
DIA
STATE
NAVY
ARMY
USAF
25X1 DDR&E
AEC
FBI
NSA
25X1 Secretary
- Dr. Bruce H. Allen
- Capt. Wendell J. Furnas
- Lt. Col., George A. Parsons
- Col. Chester H. Morneau
- Not represented
- Not represented
- Not represented
Others Present
25X1 CIA
Mr. Charles A. Briggs
25X1
CIA/NIPE
NPIC
DIA
[DIA] .
*Present for part of the meeting. 9 "R Of" rom raT Approved For Release 2005/03/16: CIA-RDP80LBOeld3ded~ra g ar8atic
owngra ng an
S-E-C-R-E-T declassification.
 Approved Fa&,Release 2005/03/16 : CIA-RDP80BO11 A000100130001-8
S-E-C-R-E-T
25X1 NSA
25X1
ARMY - Mr. John C. Wilson
CSS -
F CIA
DIA
1. A royal of Minutes. Minutes of Meeting 83 (CODIB-M-83, 2
December 19 were approved as distributed.
2. 'Computers in the Soviet Economy. 25X1
CIA/ORR, presented an unc assi ied' r e ing on this subject. A complete
text is on file in the CODIB Support Staff. Highlights are as follows:
The USSR lags badly behind most industrial nations in the use of com-
puters. The US has 10 times as many computers in use as the USSR
(35,000 in the US against 3,500 in the USSR). The major reason for the
Soviet lag is that most of the USSR computers are used in military and
scientific work and the USSR has only scratched the surface in economics
applications. Fewer than 500 outmoded computers are used in the Soviet
economic data system, whereas there are about 20,000 in use in the US
economic system. Soviet planners hope to replace their present cumber-
some methods (of collecting, compiling, and analyzing economic data) with
a national network of computers. They face major problems in achieving
this goal. They must install several thousand modern computers, train
thousands of computer specialists., develop an elaborate communication
system, revise their economic reporting system, and educate enterprise
managers in the application of computers. The URAL-16 is typical of the
Soviet computers used in the economy. It is capable of 80,000 instruc-
tions per second with a storage capacity of 130,000 words. Their best
machine is the BESM-6, capable of one million instructions per second,
but was designed for scientific rather than economic uses. The Soviet
Five Year Plan for, 1966-70 gives major emphasis to increasing production
of computers able to handle large amounts of economic data. At the some
time, Soviet economists are pushing the development of mathematical models
of the economy. It will be well into the 1970's before the national
network of computers can begin to relieve the burden of paper work and to
help solve problems for plant managers.
3. COINS Status Briefings. In response to agreement at the CODIB
meeting on 1 0 November pare 12, CODIB-M-82, 17 November 1966) for a
full CODIB session devoted to COINS, five briefings were presented as
follows:
Background and problem areas -
25X1
Support Staff
Communications Switch -
DIA 25X1
25X1 Airfield Files - DIA
Simulation of Communications Switch - 25X1
Biographies Files - NSA 25X1
25X1
COINS Eva .or uai-a eionase
Approved
981139A000100130001-8
 Approved FJelease 2005/03/16: CIA-RDP80BOl A000100130001-8
S?E-C-R-E-T
25X1 The complete text of I briefing and significant highlights
of the other briefings are given in the attachment hereto. Due to a
malfunction of the tape recorder, the complete texts of the other
briefings are not available.
25X1
It. Chairman's es onse. The Chairman thanked the briefers for
their informative -presentations. He observed that CODIB had intentionally
not given specific terms of reference to the COINS Working Group so as not
to prescribe its efforts too closely. Indeed, one option was to permit
NSA to deal bilaterally with each agency without USIB participation
directly or through. CODIB. With respect.-to COINS.II, he noted that CODIB
had expressed the feeling previously that-.-It was. premature to plan COINS
II in detail and to seek approval of extensive funding until COINS I is`
operational and has been evaluated. The-.CODIB members have indicated
that resources will, not be forthcoming until it-is demonstrated that
COINS I is a useful system. lie then asked F- _j Chairman of the.
COINS Working Group, if there was anything he needs now or later from
CODIB in order to keep progressing toward.the implementation of COINS I.
The response was negative..
5. Amount of Detail Reported in CODIB Minutes. The Chairman com-
plimented the Secretariat on the detailed minutes of previous meetings
but wondered whether such detail was helpful to the CODIB members,
Several members stated that they appreciated the detailed minutes and
found them very useful.
6. Revision to DCID 1/7. The Chairman noted that final USIB action
on the COD B-prop so drrevisions to DCID 1/7 (USIB-D-5.116, 23~November
1966; CODIB-D-77/3, 21 November 1966) awaited signature of DCI.
7. Status of Task Team III Re or . The Chairman noted that a DCI
statement of no concurrence on Recommendation 6 in the CODIB Report on
Task Team III had been sent: to USIB members (Memorandum for Holders of
USIB-D-39.7/19, 6 December 1966). [Subsequent to this CODIB meeting the
USIB Secretariat reported that on 12 December 1966 it was recorded that
the USIB approved CODIB Recommendations 1 through 5 but did not approve
Recommendation 6.)
25X1
8. National Indications Center (NIC) Request to Join COINS Network.
The Chairman tecra letter dated 6 December 1966 from the Director NIC
requesting that the NIC be considered for participation in COINS I. He
proposed CODIB approval of the request. wondered what files
25X1
proposed for COINS could possibly be of interest to NIC. 25X1
indicated that he had had preliminary discussion with representatives of
141C and that they had indicated a need for files containing some of the
same kind of information contained in the proposed COINS files. The NIC
request was then approved, and the Chairman stated that he would notify
the Director NIC suggesting that he contact of NSA to firm up 25X1
the nature and extent of NIC participation and 25X1
of DIA for assistance in obtaining the required resources.
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
S-E-C-R-E-T
 Approved Fgelease 2005/03/16 : CIA-RDP80B01A000100130001-8
10. Meeting of the Society for General S stems Research. The
Chairman s~~ute copies. of the program of the .twe t annual meeting
of the Society to be held at the Washington Hilton, December 26-30, and
suggested that CODIB should'give more attention to this type of meeting.
11. Future CODIB Meetings. It was agreed that insofar as feasible.
future C .B meetings will be held on Tuesdays at 10:00 A.M. It was.'
agreed to hold the next meeting on Tuesday, 10 January 1967. Main Agenda
items will be CODIB reports to USIB on Training and Biographies.
25X1
Attachment : A/S
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 Approved Fc telease 2005/03/16 : CIA-RDP80B011 1000100130001-8
CODIB-M-83
Attachment
25X1
25X1
COINS BRIEFING FOR CODIB
14 December 1966)
1. INTRODUCTORY NOTE
A. This briefing is a direct response to CODIB's request of
3 November 1966 for a session devoted to a discussion of COINS, its status
and its problems, in order to find out clearly where we are now, what it
is t.iat we are implementing, and to dispel existing confusion between
COINS I and COINS II. (Chart 1)
B. In the time available, it is impossible to cover all the de-
tails associated with the COINS experiment because too much has happened
during the last sixteen months. The best I can hope to do is to pre-
sent a selection of information which will form a coherent picture of
COINS at a predetermined level of abstraction. Those who will brief
after me, and the chairmen of his working panels, will provide
information at a more detailed level in some specific areas of COINS
development.
C. During this briefing I shall refer to group as 25X1
tie COINS Committee, CODIB established the group as the CODIB Working
on PFIAB Recommendation Two. The shorter title has the obvious
advantage of brevity,
II. ORIGIN
A. PFIAB Recommendation No. 2 of 15 June 1965.
1. The COINS experiment was initiated by Recommendation No. 2
of 15 June 1965 by the President's Foreign Intelligence Advisory Board
(PFIAB), The recommendation was rooted, at least in part, in a briefing
NSA gave to the Communications Panel of the PFIAB regarding the NSA
25X1 I IPS project. Next, the recommendation appears to reflect some of
the ideas contained in various Government reports on the management and
uses of AD? in the Federal Government. Finally, the recommendation seems
to be based on the clear, recognized need for improvements in the
intelligence community's capabilities for handling its information base,
2. The PFIAB recommendation called for an expansion of the NSA
TIPS project to include participation by other members of the intelligence
community as a first step toward inter-agency and inter-building informa-
tion handling. The PFIAB stated that there should be a capability for
extensive handling of the Russian biography problem by the summer of 1966,
and a capability to exchange outputs from various sources by the summer
of 1967. (Chart 2)
Group 1
Excluded from automatic
Approved For Release 2895 r$ETi4-RDP80B011 ZWt41 6810( 3'hA
declassification.
 Approved FoQtelease 2005/03/16 : CIA-RDP80B011 W 000100130001-8
S-E-C-R-E-T
1. The PFIAB provided an explanatory statement immediately
following its recommendation. For our purposes here we can view this
statement as the PFIAB guidance for carrying-out the COINS experiment.
This guidance encompassed the following points: (Chart 3)
a. Experimental trials are needed to come to grips with
a wide variety of problems involved in intelligence information handling,
including
Security Compartmentation
Encryption of communications between the
computer/information base and user locations
Other problems
b. It may be necessary to expand the scope of the infor-
mation in TIPS: however, this should be done with caution as to the total
amount of material thus added.
c. The community's intention should be to establish a
system that will in fact be used by workers in at least a few agencies as
a better way to meet day-to-day tasks.
d,. The system should be regarded as experimental and there
should be no attempt to insure that in its experimental form its opera-
tions can be economically justified.
C. PFIAB Philosophy.
1. The PFIAB expressed its views on the community's informa-
tion handling problems in some ten or twelve paragraphs praceeding the
actual recommendation and the guidance statement I have just covered.
We might view these paragraphs as constituting the PFIAB philosophy re-
garding the uses of ADP in handling intelligence information. I have
selected some of those statements as being particularly germane to this
presentation on COINS. (Chart 4)
a. Information handling methods occupy a pervasive
position in the whole administrative frame-work of the intelligence
community. They are a determining factor in the effectiveness of the
intelligence system In meeting national security needs at policy and
command levels of the Government.
b, Systems problems of intelligence information access
will continue to be of the most difficult type, heightening the impor-
tance of great improvements in the depth of understanding and of skills
in tackling a wide variety of problems which confront all levels of
Government personnel concerned with access to the national intelligence
base.
Approved For Release ?Q A5 A ' ;. lA-RDP80B01139A000100130001-8
 Approved Fo, elease 2005/03/16 : CIA-RDP80BO114A000100130001-8
c. Additional actions must go forward concurrently with
those already approved by USIB. These additional actions provide the
only foreseeable means of extending to the massive operations of the
intelligence community the advantages of high-speed machine processing
in a way which has already been applied in such specific areas as
cryptanalysis. Unless such actions are undertaken, there is danger that
the efficiency of production and dissemination of intelligence will
decline progressively, and that already high costs will climb so steeply
as to jeopardize national support of the broad intelligence effort.
d. The need for new actions is not regarded as a direct
consequence of the rise of the electronic computer. The need is more
deeply the result of growth of the intelligence community effort and the
greater growth of the information which it must handle.
e. Mechanized and automated means of access to many sorts
of intelligence cannot meet simultaneously the requirements for high
recall and the rigid requirements as to relevance. For some time to
come, the mode of gaining access to intelligence information will be
through combined machine-human systems In which the machine will retrieve
stored information in order that its relevance may be established by
human examination..''It is this combined machine-human factor which
generates systems problems of great difficulty and dimension.
SPECIAL NOTE. It should be noted:here that COMB has not provided
to the COINS a ort any guidance and philosophy corresponding to that
provided by the PFIAB. CODIB has had COINS as a major topic of dis-
cussion at ten of its meetings during the period August 1965 - December
1966. Although individual members have': expressed doubts, reservations
and concern about COINS, CODIB has not issued any substantive instructions
to tae COINS Committee. CODIB has not issued terms of reference for
COINS as it did for some nine or so Task Teams that have operated over
the past couple of years, The closest CODIB came to providing guidance
was its rejection of the COINS II portion of the Implementation Plan of
May 1966. Basically, therefore, COINS has been and still is largely the
result of bi-lateral efforts coordinated by the COINS Committee and its
five panels without direct intervention'by CODIB.
It should also be noted that there is no evidence of any sustained,
substantive overt attempt to establish communication between CODIB, the
intelligence community and the PFIAB in order to explore the real meaning
of the PFIAB recommendation and to report progress. This is difficult to
comprehend because COINS appears to be the culmination of many tough
community information handling problems, making the COINS experiment one
of the most important yet undertaken.
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 Approved Felease 2005/03/16: CIA-RDP80B011?000100130001-8
S-E-C-R-E-T
III. WHAT IS COINS
A. Introductory Note.
For the next few minutes I will address the question WHAT IS IT
WE ARE IMPLEMENTING? One cannot adequately answer the question of what
COINS is now without first considering what COINS started out to be and
how much, if any, the original concept has changed over the months. This
discussion will be introduced, therefore, by a brief reconstruction of
the early approaches to the COINS experiment.
B. Initial System Approaches.
1. At its first meeting in mid-September 1965, the COINS
Committee discussed its ultimate goal as being the "preparation of guide-
lines for an intelligence network of self-scheduling remote-access com-
puter systems communicating with one another via secure communications,
rather than a single computer system involving aset of centralized
information files". The COINS Committee acknowledged that the ingelli
gene. network of self-scheduling computers was a long range goal and
that "some of the managerial problems involved in remote interrogation
of shared information files could be faced in the near future by using'.
the TIPS PILOT system at NSA to interrogate a small sub-set of intelli:
gence files of community interest". Ue`;can see here in the beginning a.
definite ttvo-phased approach to the problem which is not at all unlike
the approach indicated by the PFIAB. In'this approach the initial effort
would1be to experiment and gain knowledge. through using the NSA TIPS
PILOT'with a small sub-set of specially'selected.intelligence files and
a number of remote interrogation stations. This effort would be geared
to establishing a knowledge.and experience base on which to build a
follow-on system of considerably more sophistication.
2. This approach continued for about a month at which time
the DIA and CIA members of the COINS Committee expressed concern that
the COINS effort would not work effectively if the DIA and CIA outstations
were located in the NSA Operations Building. Such an arrangement would:
be unnatural for the analyst user and would not be responsive to that
portion of the PFIAB recommendation concerned with the inter-building
handling of intelligence information. An apparent solution to the pro-
blem would have been to locate the remote interrogation stations in the
headquarters building of each participating agency and connect them to
NSA TIPS PILOT via secure data links. This approach was nullified when
NSA announced its policy that all COINS. outstations tied to TIPS PILOT?
would have to be physically located in the NSA Operations Building. NSA
stated that this constraint was necessary to safeguard the sensitive
information in TIPS. which had nothing to do with COINS but which could
not be effectively isolated from COINS. This NSA policy precipitated a
search for an alternate approach to the initial COINS effort. Such an
S-E-C-R-E-T
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 This chart (Chart 5) depicts schematically COINS I as it
has emerged from the search for an alternative to locating all COINS
outstations in the NSA Operations Building and conducting experiments
using.TIPS as the basis for.'design of the ultimate "intelligence network
of self-scheduling computer systems". There are a few important points"
to observe on this chart. The first is:the obvious fact that COINS I
is not simply an expansion of the NSA TIPS; TIPS is merely the internal
vehicle through which NSA ties into the COINS network and, in this regard,
appears to produce no greater impact on:?the overall COINS I experiment:
than does the CIA or DIA local computer -'systems, Related to this is the
second point that the computer systems in COINS I vary greatly, i. e.
approach was found when the DIA member of.the COINS Committee suggested
25X1 that the DIA 0 remote access computer system be interfaced with.
COINS PILOT (or COINS I as the effort was later called).
C. The Hardware System.
25X1
25X1
25X1
The third point tonote is that the COINS ar ware;:
system is dominated by the information transfer'sub-system, i.e., the
Approved Fo elease 2005/03/16: CIA-RDP80B011, 000100130001-8
This chart (Chart 6) lists the intelligence information
files ;which have been conditionally nominated for use in the COINS I
experiment. The list is shown here only to complete the picture of what
COINS is; the files will be discussed in detail later when we talk about
the status of COINS and some of the problems associated with malting COINS
a meaningful experiment.
E. System Operation.
1. I have included a brief section on system operation to
round-out the description of what COINS is, and also to provide a lead-in
tol more detailed description of the store and forward communi-
cations switch now being built at DIA. I will discuss three facets of
COINS operation: (Chart 7)
The User Analyst
The Local Computer System
The Information Transfer System
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 Approved Fa elease 2005/03/16 : CIA-RDP80B0l A000100130001-8
2. User Analyst Interrogation Operations.
a. COINS I users will be limited to file-oriented interro-
gations (as opposed to subject-oriented interrogations). This procedure
requires that the analyst have a detailed knowledge of the files he
intends to interrogate. For each file he will use, he must know the
following:
Exact name of the file
Name of the fields in the file
Data elements (how information is recorded
into the file)
In addition, the user analyst must know:how to format his request so that
it will be accepted in the: system and processed against the appropriate
file. In this regard, the'user can use either a specific or canned
interrogation or a query in a users language.
b. When using a specific or canned interrogation the
analyst is simply using an interrogation which the owner of a file has
25X1 already developed for his own operations using the file.
If the specific or canned interrogations available for
eaca COINS file do not satisfy the needs of a user in another agency, the
user has two options:
(1) Write a specific interrogation, using the language
of the agency owning the file, and have the interrogation included in the
program library of the retrieval system in the agency having the file to
be interrogated.
(2) Use whatever specific or canned interrogations are
available and write a specific program for his own local computer system
to merge and reformat the results received from other computer systems.
c. The user analyst has the job, tlien, of preparing the
query in the appropriate format and entering it into his remote console.
In COINS I, this requires that the analyst have a detailed knowledge of
both the files and the procedures for retrieving information from those
files.
3. The Local Computer System.
a. The computer systems in COINS I serve two basic roles.
First, each of the computers has as its primary role the support of the
owning agency in areas unrelated to COINS. Second, each of the computers
serves the secondary role of processing COINS interrogations and responses.
When operating in the secondary COINS role, the local computer performs
three basic Approve o'r Release 2005/03/16 : CIA-RDP80B0l139A000100130001-8
 Approved Fo elease 2005/03/16 : CIA-RDP80B0114000100130001-8
S -E-C-R?E-T
(1) It receives and processes interrogations from its
attached remote stations and prepares them for Input to the information
transfer sub-system of the COINS network.
(2) It receives responses to interrogations and routes
them to the requesting analyst.
other agencies.
(3) It receives and processes COINS interrogations from
b. When operating in the first mode, i.e., receiving and
processing interrogations from its attached remote stations, the local
computer system will perform one of two types of operations depending
upon its location.
c. When processing responses to its own interrogations,
the local computer system will either pass the responses directly back to
the requestor's station as soon as they are received, or input them to a
special program for, further processing (merging, summarizing, formatting)
before forwarding the responses to the reLluestor.
d. When receiving and processing interrogations from other
agencies, the local. computer system will perform four basic functions:
(1) receive incoming interrogations and send back
receipt messages
(2) validate each interrogation by determining that
the specific or canned interrogation and the appropriate data files are
available and that the requesting agency is authorized to use the programs
and files
('3) respond with service messages if the interrogation
is in error or unacceptable
('IS) process the interrogation against the appropriate
file, and prepare and forward a proper response to the requesting agency.
4. The Information Transfer System.
a. The information transfer system is a major portion of
the COINS experiment as it is now constituted. The heart of the transfer
Approved For Release 2855491W:-CIA-RDP80B01139A000100130001-8
 Approved Fo elease 2005/03/16 : CIA-RDP80BO11 000100130001-8
system is the store and forward communications switch built around the
25X1 computer. will cover the details of the switch later.
It is sufficient for our purpose to note only that the switch is the
unit which drives the system and maintains order within it.
~~. It should be noted that the total transfer system
(charts 5 and 7) is comprised of a relatively large number of individual
units which must all function properly to insure that information can be
exchanged securely within COINS. High reliability in such a network is
difficult to achieve. Consider for a moment the way system reliability
is calculated. Assume we have a network of three individual units (U1,
U2, and U3) in series, each with its own reliability factor (Ri, R2, and
R3) and each equally dependent on the other for total system performance.
The reliability of the total three-unit system is not that of the indivi-
dual units but rather is the product of R1 x R2 x R3. If Rl m .9, R2 : .9
and R3 - .9, the reliability of the total system would be .729.
. The factors discussed above make the COINS information
transfer something more difficult than simply connecting together a
number of black boxes and throwing an actuating switch.
_y. WHAT IS THE STATUS OF COINS
A. Introductory Note,
Although testing of the COINS communications link between
DIA and NSA is scheduled to get underway on 21 December 1966, the entire
COINS network will not be ready to commence operation until at least
next summer if the present rate of development continues to hold. In
:.is last report to CODIB, the Chairman of the COINS Committee said the
operational target date might be as late as 1 July 1967. Subsequent
conversations with personnel working on COINS have indicated that the
operational date may have slipped beyond 1 July.
This chart (Chart 8) is designed to show the overall status
of COINS by component and participating agency. The chart is not com-
pletely satisfactory because it is difficult to assign uniform, positive
values to all the components. Almost every block shown on the chart
warrants some degree of elaboration to establish precise status. For
e,smple, I have shown the DIA non-COINS software as "basically ready`?; in
25X1 actuality, the DIA I system is operational but there are still
some parts of the system to be brought on-line.
25X1
Computer Hardware (Local Computer Systems) is ready.
Approved For Release 20 Q.3R:6E42A-RDP80B01139A000100130001-8
 Approved Fo Release 2005/03/16 : CIA-RDP80B01tW000100130001-8
S-E-C-R-E?T
$. Non-COINS Software (that software which each agency
-ould develop for its own local computer system even if there was no
COINS experiment) is basically ready in CIA and DIA. NSA's software for
L:,ggraded _rZM is not ready because of what described
'{is last status briefing as delays in solving the multi-level security
i?o npa rtmentation problem. NPIC's status is about the same as NSA's
#PL!atuse NPIC will use at least some of the NSA software.
COINS Software (that software to be developed by each
df uey specifically for participation In COINS) is basically not ready.
In the case of NSA, the COINS software is dependent upon the ["PIPS
,ortwsre and will necessarily lag development here. All participants have
Jorwe some deeirm and coding of COINS software and are into early testing.
4. The COINS files are shown as not ready. This means
that the files are not ready for immediate use If one wanted to commence
operation today. The files are largely in machinable form but still
}ave to be worked on to make them suitable for COINS use. It has been
estimated that approximately one month of work would be required to get
some of the files in shape for actual COINS usage.
S. The Remote Terminals are basically ready. In the case
of State Department', approval for acquisition has been given and procure-
ment is all that remains to be done.
6. Crypto Devices and MODEMS are basically ready.
7. The Communications Switch Hardware is installed.
B. The Communications Switch Software has been largely
designed and is being developed by DIA, NSA and
0
9. COINS Training is shown as not ready because the various
users at each agency have not been trained in the files and interrogation
procedures of the participants, This training is being scheduled and
will be completed within the framework of the summer-1967 operational
target date.
10. COINS Evaluation planning is not ready. The technical
evaluation plan is in an advanced state; however, the operational evalua.
tion plan remains an unsolved, difficult problem.
11. COINS Objectives are still a major problem and will be
discussed later in the briefing.
C. The COINS Files.
1. The selection of suitable files for the COINS experiment
ras been (and still is) a most difficult problem. The file selection
problem can be approached, at least theoretically, in two ways:
25X
25X1
25X1
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
S-E-C-R-E-T
 Approved Foa?itelease 2005/03/16 : CIA-RDP80B0114000100130001-8
S-E-C-R-E-T
- 10 -
a. Define the problem area(s) against which COINS is to .
be tested and, then, select or develop the files required for effective;:
treatment of the problem.
b. Use whatever files are available and, then, manufacture
a problem area where the files can be used.
2. The first approach is obviously the more logical; however,
the second approach has some appealing aspects under conditions of
limited time and resources for building new files. As COIDB reported to
USIB in September 1966--"the real difficulty is in selecting files which
are readily available in machinable form and which, at the same time,..,'
nave high utility to the analysts who will participate in the operational
feasibility portions of the experiment".'
3. This chart (Chart 9) lists the files which have been pro-
visionally scheduled for use in the COINS experiment. Also shown is the
approximate number of records or characters in each file, and the
estimated rate of growth of each file.
. Following is a brief description of the contents of each
25X1 file and, in some cases, an indication what kinds of services can be
obtained from the file:
Approved For Release 2005/03/16 : CIA-RDP80B0l139A000100130001-8
 25X1 Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
Next 2 Page(s) In Document Exempt
Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
 Approved Felease 2005/03/16: CIA-RDP80B01A000100130001-8
S-E-C-R-E-T
- 14 -
V. WILT IS THE DIFFERENCE BETWEEN COINS I AND COINS II?
A. Introductory Note.
In a way it is rather awkward to talk about the differences
between COINS I and COINS II because COh)IB disengaged itself from the
COINS II portion of the COINS Implementation Plan of 25 May 1966. This
disengagement may have been less than complete, however, because several
of the activities of the COINS Committee are of a nature that suggest
the original COINS II concept is still very much alive. Using the COINS
Implementation Plan as a point of reference, therefore, some of the
differences between COINS I'and COINS II are enumerated in the following
section. (Chart 10)
25X1
C. Intelligence Files.
The COINS I files are relatively non-standard and of
questionable efficiency and utility. An objective of COINS II is to have
files which have greater standardization, efficiency and utility.
1. COINS I uses a mixture of interrogation and communications
languages; COINS II will strive for common interrogation and communications
languages throughout the network.
Approved For Release 2005/03/16 : CIA-RDP80B0t139A000100130001-8
 Approved Fg6Release 2005/03/16 : CIA-RDP80B011, 4000100130001-8
S-E-C-R-E-T
. 15 -
2 In COINS I all information, regardless of actual security
classification, will be protected to the Top Secret SI level; COINS II
will have the capability to handle multi-level security classification.
3. COINS I is limited to file-oriented interrogations; COINS
11 will have a subject-oriented interrogation capability which will not
require the user to know anything about the file he is interrogating.
VI. WHAT ARE THE PROBLEMS IN COINS I?
A. Introductory Note.
The identification of problems in COINS I is a somewhat
subjective process dependent upon the particular bias of the individual
making the identification. In developing a list of problems, one can't
get much assistance from reviewing the written material available on
COINS because the participants, although they have complained about
aspects of COINS, have not extended themselves very far in setting down
their hard, substantive objections to COINS I. Accordingly, I have a
list of problems which are particularly appealing to my own bias and
whic.i deal largely with policy, rather than technical, matters.
B. Management.
There has been no effective CODIB/USIB management of the
COINS effort since its initiation. COINS Committee must be
acknowledged, on the basis of past performance, as the only central
management influence in COINS. Except for its review of the COINS
Implementation Plan and a few general briefings scattered over the last
sixteen months, COMB has not maintained management contact with the
system design and development effort. CODIB, in its COINS Progress
Report to USIB in September 1966, stated that there is a need for a more
formal management structure--one that is empowered to exercise broader
authority over the entire network and its impact on the overall functions
of the community. Related to the COINS management problem is the fact
that CODIB has not provided substantive guidance to the COINS effort.
C. COINS Objectives.
1. COINS technical objectives seem to be fairly well set and
center mainly around the performance of the information transfer system.
The picture is not so bright with respect to COINS operational objectives:
approved operational objectives simply don't exist.
2. This chart (Chart 11) shows some of the evolutions through
which COINS objectives have passed. The PPIAB seemed to view the objective
25X1
Approved For Release 2005/03/16 : CIA-RDP80B0l139A000100130001-8
 Approved Fo lease 2005/03/16 : CIA-RDP80B0119 W0001 00130001-
as an experiment in the inter-agency and inter-building exchange of
intelligence information using mechanized methods, with the specific
goals of being able to handle Russian biographies by summer-1966 and to
exchange outputs from various sources by the summer-1967. In its very
early work, the COINS Committee stated that a primary objective of COINS
was solution of the problems of organizational parochialism.
3. The COINS Implementation Plan listed the following as the
ultimate objectives of COINS:
a. Reduce duplication of effort by eliminating necessity
for maintaining and supporting multiplicity of EDP programs and formatted
files of similar content by direct inter-agency computer communications.
b. Improve the community's capability to exploit the
ever-increasing volume of intelligence by improving timeliness in pro-
cessing, maintenance and distribution of finished, semi-finished, and
hey intelligence information.
c. Provide a high degree of flexibility in managing,
selecting, collating and distributing intelligence information.
d. Improve the opportunity for the effective utilization
of finished, semi-finished and key intelligence information by making it
readily accessible to technicians at various consumer and intelligence
producing agencies in a useful time frame.
e. Establish a basis for designing and constructing a more
sophisticated, dynamic intelligence network in the future.
f. Provide for more effective and efficient utilization of
equipment, manpower and time.
g. Develop the security requirements and controls necessary
for dynamic intelligence information exchange.
L. In Its progress report to USIB, CODIB stated that it had
commenced tie preparation of a list of COINS objectives together with an
itemized break-out of both technical, and operational performance
specifications.
5. In view of the questions left unanswered by the foregoing.
I have looked into the work of Biographies Panel to deter-
mine whether the work in a particular subject area of COINS might give a
better insight into what COINS hopes to accomplish. This review did
provide a little more information.
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 Approved For%Release 2005/03/16: CIA-RDP80BO11 ,#000100130001-8
S -E-C-R-E-T
- 17 -
6. The minutes of the COINS/Soviet Personality Panel meeting
of 26 September 1966 states that the PFIAB had stressed the importance
of biographies and, therefore, that a major objective of COINS Is to
develop a long range program for handling and mechanizing, wherever
practical, t.%e Soviet biographic and personality files in the community.
(Chart 12) The Panel noted that the problem is very large, very complex,
and very parochial. The Panel concluded that the approach to the problem
should be a detailed study of four areas (i.e., Sources, Collection,
Files, and Users and Their Requirements) using a small sub-set of files
in the COINS experiment. The actions to be taken in each of the four,
areas are shown below: (Chart 13)
a. SOURCES
Determine extent of duplication and, if excessive,
develop a plan to eliminate unnecessary duplication
In the future
Determine if there are any gaps in sources of
information and,if there is, develop a plan to
cover it as quickly as possible.
b. COLLECTION
Study and identify the responsibilities of each
agency and determine whether these responsibilities
are being adequately met. If not, make
recommendations to see that they are met.
Examine collection requirements and procedures for
each source and indicate the amount of control
that can be exercised by the community.
Develop techniques for collecting information direct
from the source in machineable form for direct
input to machine: systems.
Eliminate unnecessary duplication in the collection
and processing of information; also, eliminate
cumbersome and time consuming collection techniques.
c a FILES
Files are essentially of two types--those which lend
themselves to machine processing and those which
do not. This study must develop procedures for
adequately handling both types of files.
d. USERS AND THEIR REQUIREMENTS
Answer such questions as:
Why are these files being established and maintained?
S-E-C-R-E-T
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 Approved Folease 2005/03/16 : CIA-RDP80B011 1000100130001-8
S-E-C-R-E-T
What files need to be established, combined or
discontinued?
What are the current and projected requirements
for Soviet biographic information?
7. The Personality Panel stated that the results of the study
of the four areas Just described would enable the COINS Committee to:
a. Uncover and identify problems involved (data standards,
name variants and variations, transliterations, etc.).
b. Develop and recommend a coordinated, community-wide set
of proposed solutions for handling each problem area.
c. Initiate a pilot community-wide program to test the
proposed solutions.,
d. Develop a set of tested procedures and mechanisms for
studying the entire Soviet biographic and personality problem in the
intelligence community.
1. The COINS files have been a major problem since the begin-
ning of the effort. The big concern is t:iat the files now nominated do
not contain high-utility information which will induce the analyst to
use COINS and, thereby, contribute to our overall knowledge of the real
values of COINS and. how to capitalize on these values in designing
effective follow-on systems. The problem with the files is connected to
the problem of COINS objectives. If specific operational objectives were
set, then it would be possible to select or develop those files best
suited to the achievement of those objectives. The files now nominated
are largely a collection of what is most readily available and not what
is most needed for a truly effective experiment.
2. The files problem introduces a range of other problems
dealing with such matters as data element standardization, file centrali-
zation and maintenance, elimination of unnecessary duplication in the
collection and input of information, and the realignment of community
responsibilities for various intelligence functions. The difficulty
here is determining which of these problems must be solved as a pre-
condition to conducting the COINS I experiment, which problems should be
set aside until COINS I has been completed, and which problems really
don't belong within the purview of the COINS Committee.
E. COINS Evaluation.
There exist no hard, tangible plans for the operational
evaluation of COINS I. Current plans (those prepared by the NSA
Approved For Release ~f l tg-bA-RDP80B01139A000100130001-8
 Approved Fo elease 2005/03/16: CIA-RDP80B0114 000100130001-8
S-E-C-R-E-T
- 19 -
Operations Researcih, Staff) cover only the. technical evaluation and will
probably produce a fine set.of quantitative data on the performance of
the information transfer system. These data by themselves, however,
probably would add little to what we already suspect about COINS--that
is, a network of reasonably effective communications equipment and
electronics can move information from one point to another faster than
the same information can be moved by existing manual methods. These same
quantitative data would have to be squeezed very hard, indeed, to yield
useful insights into the value of COINS as an analyst support tool.
Approved For Release5205 301 j : f IA-RDP80B01139A000100130001-8
 Approved ForRelease 2005/03/16: CIA-RDP80B011.WA000100130001-8
e .eeeion of cons
its stet=
its problem
'To Ind out
ihere we are mm
,what it in we fan: 3.r oIea Anti2w
impel emoting eon tue about the diffe ez bete OOM I armd WVIM U
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 Approved Foo-Release 2005/03/16: CIA-RDP80B011-&&A000100130001-8
F 71 o Prot
ipation by other agamiea
- . , tel operating systm
lnt+nr-sgamy
later..--bulY filing
Informtiori bsadlio1
ftte irs band'Ling of Rue. sn biogmpby - maser 1966
car outputs f c variota scum $ - sLVOer 1967
Opsrati1) .1 Mda"
S-vmzri-t;V C s rLumatation
:won of iioatioos
Cither IProbless
y RRvs to Rm=rd & opo of Infor atios in TUB
Used for Day-'t o-Dgy Work
Attompt to Inure 'Zbat OpIrstion of cperioentm1 System call Be
Ay Justtfi..ed
Approved For Relea9e'20 5f'&JA 6: CIA-RDP80B0l139A000100130001-8
 Approved Foelease 2005/03/16 : CIA-RDP80B014.MA000100130001-8
l*srvariVaaese of ; nfoz tion Hana1t )t&
i*portsr a of set tavo~-oacts
vautagss of NLdiirie Processing
end 19cperimee in Cryptanalysis/c xmmicastiom
3l ause of )4h4 systMW
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 25X1 Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
Next 2 Page(s) In Document Exempt
Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
 Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
~ ~ p
e
t
' n-mir R1!re
wot- Ra ikv
- ft-miall"'r m
rnm4 irn
ftt near
/A
AIP
!:'DID Softie
Dsst~s
Ds.t
Dssigm
ILA
coin Fugw
lot
lot
!
Dae
&/A
~.
gate ?Si'$ ar1S
D_
DY
Dad N"Y
lt*
MA um
aK
ft"y
Not
ftitah Barb are
&A
C
i/A
I/A
N/A
MA Mm"
?A
A
A
nut"m
~o+t
fot Assay
lot
lot
dot Res+dy
481AC!'IY'A'e
SAL TICr
CLmML - lot Away
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 25X1 Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
Next 1 Page(s) In Document Exempt
Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
 B-Z-C-$-$-T
Cl art U
coosq m
i s bas oomssaced list of COM objectives
20 October 1965
Approved F elease 2005/03/16 : CIA-RDP80B0I 000100130001-8
Solve the pzobleo of or niztioval parochLLnlism.
25 Sky 1966
Reduce duplication
Improve espicitation
Provide tlitibt1ity
prove uti3.i=atioa
1Dstab-lish base for future network
Provide better use of resources
Develop security controls
1 September 1966
Approved For Release 2005/03/16 : CIA-RDP80B0l139A000100130001-8
 Approved Faelease 2005/03/16 : CIA-RDP80B01'i 000100130001-8
COMB BIC APSTCS
A 1/W1Cgt OBi CTIVE
Develop a long range program for Handling Sad mechanising, wherever
practical,, Soviet biographic and perso1lity files in the
intelligence cc unity.
ADPAOI R UIR STWI07
Coiection
responsibilities
techniques
Ike= Sad Requirements
Approved For Release 2005/03/16 : CIA-RDP80B0l139A000100130001-8
 Approved Fc elease 2005/03/16 : CIA-RDP80BO11W000100130001-8
Mart 13
cons COMMUM TO -
Identify Prot ess
Data Standards
fume Variants and Variations
Transliteration
etc.
Develop Coordinated Solutions
Initiate Cammsuity-vide Test
Develop Set of Tested Procedures and kacbanisms
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
 25X1 Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
Next 4 Page(s) In Document Exempt
Approved For Release 2005/03/16 : CIA-RDP80B01139A000100130001-8
 Approved FgRelease 2005/03/16 : CIA-RDP80B01V,A000100130001-8
COINS EVALUATION
'Highlights of Briefing byl DIA) 25X1
COMB, in its progress report to USIB in September 1966, stated the
need to further consider what we are embarked on in COINS I. The subiect
of COINS evaluation has been discussed several times in the COINS Working
Group; a COINS Evaluation Panel was established in November 1966. The
Panel has met twice. The material to be discussed here is the result of
two panel meetings.
There are two types of evaluation to be conducted on COINS I: the
Technical Evaluation which we can see how to do, and the Operational
Evaluation which we don't yet know how to perform.
The approach to evaluation is as follows:
a? Evaluation of the Transmission System. Here, we will evaluate
the actual performance o the information transfer system to determine
how it operated. We will compare this performance with alternative
methods of transmitting information (e.g., courier, tape exchange),
Papers dealing with these two aspects of evaluation of the transmission
system are now in draft form.
b. Utilization and Acceptance. The effort here will be to
determine who is using the system and to get a general measure of its
acceptance by the user. A paper on this problem is under development.
c. Utility. The problem here is to determine how useful COINS is
in solving real intelligence problems. We feel we ought to pick a set
of problems and then decide whether COINS can be used to solve them.
The problems selected should be time-dependent, i.e., ones in which it
makes some difference how fast we transfer information.
In the evaluation of COINS utility (operational evaluation), we
have looked at the current content of the COINS files. The biographic
and airfield files offer some promise but neither of these areas is in
the high time-dependency category. We are looking into some new problem
areas in an attempt to select areas for applying COINS I. These areas 25X1
include:
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
-E-C-R-E-T
 Approved FGZRelease 2005/03/16 : CIA-RDP80BO11WA000100130001-8
25X1
The Evaluation Panel will come to CODIB with its identification of
a set of alternative problemsfor consideration.
Approved For Release 2005/03/16 : CIA-RDP80BOl139A000100130001-8
8-E-C-R-E-T