WITNO6090100
WITN06090100
Witness Name: Andrew John Dennis Simpkins
Statement No.: WITNO609_01
Exhibits: WITNO609_01/1 — WITNO609_04/18
Dated: 13'" September 2022
POST OFFICE HORIZON IT INQUIRY
FIRST WITNESS STATEMENT OF Andrew John Dennis Simpkins
This statement has been prepared in response to the Rule 9 Request of 8" June 2022
and covers the topics of Horizon planning, testing, acceptance, live trial and rollout.
Brief Professional Background
1. My professional background is in Information Technology Management and
Management Consultancy. After a brief period as a Programmer, I joined the
Management Consultancy practice of Coopers & Lybrand in 1978 where I
worked as a systems analyst and project manager on a range of IT projects
delivering retail and reinsurance accounting systems. I later undertook a
number of IT strategic planning assignments reporting to one of the IT Division
directors. I joined TSB (subsequently Lloyds TSB) in 1988 where I had senior
management roles reporting to the IT Director in the areas of internal
consultancy, programme management and IT strategic planning. In 1992 I was
the lead IT Manager on a joint IT and Finance Division project to upgrade the
Bank's financial reporting systems (reporting to the Finance Director) and was
subsequently appointed a member of the IT Division’s Management Team
responsible for financial systems reporting to the IT Director. I joined French
Thornton in 1997 as a Managing Consultant, where I worked on large-scale
1I 20
WITNO6090100
WITN06090100
public sector IT projects for the Post Office, Inland Revenue, and the
Department of the Environment. In these roles I would typically report both to a
Director on the client side and to one in French Thornton (or Serco after its
takeover of French Thornton). I was a freelance consultant between 2007 and
2013 and then took an academic role on a Masters’ leadership programme
before retiring in 2018.
2. I should note that the events I have been asked to comment on occurred 23 to
25 years ago. I have therefore had to rely on the 58 sets of documents I was
provided with to remind me of much of the detail of what took place. There are
gaps in the chronology of these documents covering my initial planning work
between June and August 1998 and for the period of the Live Trial from late
April to July 1999. Also, I only had access to some examples rather than a
continuous series of the minutes and status reports of the regular Horizon
checkpoint meetings.
Background to My Involvement in the Horizon Project
3. French Thornton provided consultancy services to Post Office Counters Limited
(POCL) and I was assigned to the Horizon Project and its Programme
Management Team in April 1998. (When I refer to the ‘Horizon Team’ below, I
mean only the POCL team). My role was focussed on Release Management
which was concerned with the negotiation and co-ordination of agreed plans
between POCL, ICL Pathway and the Benefits Agency (BA), and with the
release authorisations for the entry into Live Trial and Rollout. I was neither
2/20
WITNO6090100
WITN06090100
involved in commercial or contractual negotiations between POCL and ICL
Pathway nor in negotiations between POCL and the Government. I was aware
that there were serious commercial and contractual issues, for example from
ICL Pathway’s earlier delays in delivering the Programme, and that any
replanning of the project had to be undertaken within the terms and constraints
of the contract. For example, neither POCL nor BA had the contractual right to
tell ICL Pathway to extend a period of testing as this could be interpreted as
causing a delay. The Post Office’s key sanction was to decline the quality of the
system submitted by ICL Pathway for release authorisation into live running.
Therefore, agreement that the contractual acceptance specifications had been
satisfied prior to National Rollout, was the critical quality control step in the
Programme.
Release Management and Programme Replanning — April to September 1998
4. At the time I joined Horizon, ICL Pathway acknowledged they were working to
‘high-risk’ dates for delivery, in other words dates that were unlikely to be
achieved. It was also clear to me that there was a lack of clarity around the
complex integration of tasks required from ICL Pathway, POCL and BA to
deliver the multiple elements that were to be included in the Horizon service. I
began work on developing and negotiating a co-ordinated plan, visible to all
parties, which showed all the key steps leading up to National Rollout of the
first release (NR2), the inclusion of multiple benefits, and a subsequent release
(NR2+) which was to include the Logistics Feeder Service. The overall
3 I 20
WITNO6090100
WITN06090100
methodology of the Programme was not unusual or contentious in terms of test
cycles of both the office counter system and the end-to-end interfaces, a final
Model Office Test, followed by release authorisation for the Live Trial and then,
subject to acceptance, National Rollout. The challenge was to develop a
comprehensive and integrated plan of the required key activities of ICL
Pathway, POCL and BA, with dates that were considered achievable based on
experience to date, estimates of the work to be completed, and including
provisions for contingency. It is important to understand that the Horizon service
required not only the ICL Pathway software deliverables but integration with,
and changes to, the POCL and BA systems, a large-scale staff training
programme, and the establishment of ongoing ‘service management’ functions
comprising the users’ help desk and technical support to the live system.
. The work on planning was given extra impetus by the Treasury Review of the
Programme in the summer of 1998 due to the need to increase confidence in
deliverability. At that time, myself and others on the Horizon Team proposed a
key change in the implementation approach to reduce risk. This was to deliver
Child Benefit as the only benefit in the initial Rollout and to follow this with a
second release providing card payment of multiple benefits. This approach
decoupled the dependency between Horizon and the BA CAPS system for the
initial implementation and resulted in a more robust plan. In September 1998,
following a lengthy period of discussions and meetings, an overall integrated
plan based on this premise was accepted by all parties — the Horizon Level 1
plans (WITNO609_01/1). This plan proposed that three test cycles that had
started in August 1998 would be followed by a final Model Office and End-to-
se 4I 20
WITNO6090100
WITNO06090100
End test and complete by Christmas 1998. There would then follow a first
release authorisation, the migration and take-on of additional offices, the
commencement of the Live Trial in March 1999 and National Rollout in July
1999. The plan included contingency periods due to uncertainties over the
period of testing that would be required. The BA took a particularly sceptical
view of the timescales for Pathway testing and delivery. The Horizon Release
Management Team wrote an internal briefing note in September, Response to
Concerns (included in WITNO609_01/2) which sought to provide a balance
between BA concerns and ICL Pathway’s more optimistic outlook that testing
could be completed before Christmas. It was recognised, however, that this
date could not be guaranteed and the plan at this stage included a further four
weeks contingency for testing.
.. The motivation behind the September plan was to set out a more realistic
timescale for delivery of the Programme. It was not a response to business or
financial pressures to achieve high-risk dates. I presented the plan to the leader
of the Treasury Review and the Review's comments in October largely
supported the plan and indeed identified areas which could potentially ‘unlock’
time and bring forward the date of National Rollout. Their view on the
Programme plan and dates was included in a Presentation to CAPS Board
which I gave on 15! January 1999 (WINT0609_01/3). In September 1998,
however, we did not yet have experience of how the system would perform in
intensive testing and the live environment. It was understood that sufficient
testing time and live proving was essential if POCL was to have confidence that
the Horizon system had met the contractual acceptance criteria. The
5 I 20
WITNO6090100
WITNO06090100
Programme also put in place a schedule of meetings and checkpoints to
monitor and assess progress as the testing period unfolded.
The Testing Period August to early December 1998
7. The initial progress reports on testing during this period appeared positive with
the test cycles running to schedule and the progressive elimination of faults (or
‘PinICLs’) by ICL Pathway. Horizon was a complex system requiring hundreds
of test cases and scripts covering the counter plus the interfaces to the POCL
and BA systems. Testing cycles required the investigation, clearing and
retesting of hundreds of fault reports, not in itself unusual on a Programme of
this size.
8. In late October with the completion of the second of three test cycles, a Testing
Review was conducted, where it became clear that serious concerns had arisen
with the accounting and reconciliation processes, especially with the cash
account production in the test outlets and with the accounting results passed to
the POCL back-end system (TIP). See ‘Proving the TIP Interface - Draft Report’
(WITNO609_01/4). The completion in mid-November of the third cycle of MOT
and the second pass of End-to-End testing did not lead to a substantive
improvement in the situation. I submitted a briefing note to the Horizon Team
after completion of the third cycle on 20" November, Briefing Note on Status of
Testing 20" November 1998 (WITN0609_01/5), which concluded that ‘while
counter activity is showing progressive improvements in quality, we have not
demonstrated the end-to-end data and financial integrity of the system to the
6I 20
WITNO6090100
WITN06090100
extent that is required for entry to the final MOT and E2E run’. I summarised
the principal areas of concern as errors in cash account production, reference
data inconsistencies, errors in counter reports that prevented reconciliation at
the outlets, and report reconciliation problems with the major interfaces to the
TIP, BES, and HAPS systems.
. lt was agreed between the Horizon Team and ICL Pathway that these faults
had to be addressed to give an assurance of the quality of the system before
the start of the final Model Office Test. This meant deferring the Model Office
and End-to-End final pass of testing into late January.1999. This would in turn
result in delays to office migration, Live Trial, and Rollout with consequences
for the time of contractual acceptance. I provided a full summary of the status
of testing and its impact on the plan to the Horizon Team and ICL Pathway
management in a memo of 4" December 1998 — Horizon Testing and
Programme Plan — Current Status (WINT0609_01/6). This included the
statement on behalf of the Horizon Team that ‘Our position remains, however,
that based on the nature of the business processes involved we need to see
clear evidence of, firstly, a stable accounting and reconciliation position in the
outlet followed, secondly, by the transfer of accurate data across the
TIPS/HAPS/BES and reference Data interfaces.’ At this point in early
December 1998, there was therefore an unequivocal assessment regarding the
seriousness of faults that had been found in these first cycles of testing. ICL
Pathway accepted the need to address these faults and that additional testing
time was required.
7120
WITNO6090100
WITNO06090100
Fault Clearance and Re-Scheduling of Plans — December 1998 to January 1999
10. The situation described above led to an intensive period during December and
January where ICL Pathway sought to clear faults, and. POCL conducted ‘pre-
proving’ tests to check that faults had been cleared. These actions were
essential before entry to the final cycles of Model Office and End-to-End testing
as these would need to provide the evidence that the system was in a sufficient
state to enter the Live Trial. The additional testing activities had an impact on
the Programme plan and on the dates that had been proposed by the Treasury
Review. In summary, about 6 weeks was added to the testing time and National
Rollout was put back by about a month from late July to end August. David
Miller and myself presented a detailed report on the status of testing and the
impact on plans in a Presentation to CAPS Board on 15" January 1999
(previously referred to as WINTO609_01/3). This meeting and presentation
were part of a wider process at the time of maintaining alignment on plans with
BA and their CAPS system not only for the initial Child Benefit Rollout (NR2)
but also the concurrent activities of moving towards the testing and live trial of
multi-benefit functionality proposed for September 1999 and the development
and testing of the first main upgrade to the system (NR2+ later renamed as
CSR+) that was also to take place in 1999. All of these inter-related activities
required me to continually synchronise and agree plans between POCL, BA
and ICL Pathway during 1999.
8I 20
WITNO6090100
WITN06090100
Final Model Office and End-to-End Testing — February to March 1999 ~
11.Following the fault clearance activity by ICL Pathway, and the pre-proving
testing by Horizon staff, the final cycles of End-to-End and Model Office Testing
commenced in February 1999. I was not personally involved in the execution
or evaluation of testing, as this fell to the specialist testing teams but I was
copied on progress reports and the final evaluation reports. These reports, End
to End Testing Evaluation Report (WINT0609_ 01/7), and Model Office Testing
Evaluation Report (WINT0609_01/8) provided detailed information on these
exercises and the identification and rectification of faults. The reports do show
significant progress had been made but there were still a number of concerns.
Both cycles ran to schedule with no slippage and a large majority of previously
identified faults were reported as cleared. Some new faults were identified,
however, and these were subject to a subsequent short phase of ‘Target
Testing’ to ensure they had been corrected. Several reported incidents
remained but the assessment of the Horizon Team and wider POCL
management were that these were not critical ‘show-stoppers’ for the Live Trial
period. There were, however, concerns regarding the complexity of parts of the
system for the staff in the outlets. These concerns emphasised the need for
adequate training of staff especially in complex areas such as stock unit
balancing and the recording of losses and gains within the wider task of office
accounting. Another area identified as needing careful attention in live running
was the ongoing reconciliation of accounting data between the outlets and the
9] 20
WITNO6090100
WITN06090100
back-end systems and the accurate synchronisation of updates to the live
reference data which could impact the accuracy of reporting. These were
important observations that would need to be addressed in the Live Trial and in
National Rollout, and indeed in ongoing operation of the system.
12. By late March 1999, the Programme had therefore come to the crucial decision
point of whether to grant ‘Release Authorisation’ for the start of the Live Trial.
The Horizon Team and key POCL stakeholders were supportive of doing so but
BA claimed that the number of incidents reported in the final End-to-End and
MOT cycles required that a further cycle of tests should be carried out. The
Horizon Team disagreed with this assessment based on the limited benefit of
another test cycle compared to the value to be gained from monitoring and
evaluating the experience of actual users in a limited Live Trial. The Live Trial
would also provide vital evidence on the adequacy of training, user guides, and
the help desk. When agreement could not be reached with BA on release
authorisation, POCL decided to grant release authorisation on its own. The
earlier decision to limit the initial Rollout to include only Child benefit meant
there was no necessary dependence on changes to BA systems so this was
contractually possible. On 20" April a letter was sent from the Post Office Chief
Executive to the BA Chief Executive informing him of the decision to proceed
and urging his co-operation — Letter from Stuart Sweetman to Peter Mathison
(WINTO609_01/9).
The Live Trial Period April to July 1999
e 10I 20
WITNO6090100
WITN06090100
13.It was expected that issues would be identified in the Live Trial, which is of
course the reason for running such an exercise. I was appointed the ‘Live Trial
Manager’ for the Horizon Team following a presentation I gave to the Team on
24' February included in Horizon Management Team Meeting — Managing the
Live Trial (WINTO609_01/10). The Live Trial required the involvement of many
different units within POCL, as well as teams within ICL Pathway. My role was
not to have ‘hands-on’ involvement in the outlets but to co-ordinate the plans
and reporting for the Live Trial by the various participants, and to ensure errors
or ‘incidents’ were raised and addressed. Severity rankings were allocated to
incidents separately by POCL and ICL Pathway. These rankings often did not
agree in terms of whether they were ‘Low’, ‘Medium’ or ‘High’ which led to
extended discussions and negotiations. Cash account accuracy problems,
concerns over the adequacy of staff training to deal with this complex activity,
and the high number of resulting calls to the help desk, which were often not
dealt with quickly or easily, all featured among the ‘High’ incidents, as did
ongoing issues with the transfer of accounting data into the POCL systems.
These concerns began to extend the period of Live Trial and delay the decision
on Rollout. While faults continued to be cleared, there were still a number of
high priority incidents outstanding in August 1999, as shown in the summary at
13' August 1999 in the Acceptance Incident Hot List (WINTO609_01/11).
14.The authorisation and acceptance process leading to National Rollout had
commenced before the formal end of the Live Trial as per the contract. Given
the contractual terms around acceptance of the system this became a complex
and detailed process which required POCL and ICL Pathway to agree the
© 11] 20
WITNO6090100
WITN06090100
severity and action required on all outstanding incidents, and whether incidents
had been satisfactorily rectified and closed. A series of meetings over several
weeks monitored the evidence of the Live Trial and debated the status of
incidents and progress, or lack of it, towards acceptance. The minutes of one
Management Resolution Meeting on 12'* August (WINTO609_01/12) gives
examples of the analysis and actions proposed on incidents. I was aware that
there were also substantive legal and commercial issues to be managed at this
stage although I was not personally involved in these. Difficulties in reaching
agreement on the severity and rectification of incidents began to delay National
Rollout from the target date of 23° August 1999. Release authorisation for
National Rollout was deferred to end September but it was agreed to add more
Offices to the Live Trial. This was seen as helping to de-risk the authorisation
and acceptance decisions and the Rollout by giving further evidence of live
performance and of the rollout processes, as explained in ‘Acceptance —
Additional Offices’ (WINTO609_01/13). I continued to revise the overall
Programme plan to reflect these changes and to help assess their implications
for Rollout, and the subsequent phases of multi-benefit and the NR2+ release.
The Delay over Acceptance, Release Authorisation and the Start of National
Rollout — August to December 1999
15. During this period the live offices were monitored for ongoing problems in such
key areas of system stability (‘lockouts’ or screen freezes), cash account
discrepancies, and the performance of the help desk with regard to resolving
12 I 20
WITNO6090100
WITN06090100
calls from outlets, particularly regarding cash accounts. I attended several
progress meetings on acceptance during August, for example on 315 August
1999 — as minuted in Acceptance Update 4 (WINTO609_01/14). I provided
input, in the light of the delays, on how the transition to Rollout could best be
managed in the period running up to Christmas 1999 and into January 2000 as
for example in a memo 1999 Rollout I issued on 14‘ September
(WINT0609_01/15). I was not involved in the finalisation of the agreement on
acceptance and in the commitment to Rollout although I was kept informed of
progress in case of the potential need for replanning if Rollout was further
postponed, as noted for example in the Extract from Rollout-out Decision
included in the papers for the Roll-out 2000 Pre-Checkpoint Meeting on 15
November 1999 (WINT0609_ 01/16).
My Change of Role to the CSR+ Project
16.With the transition to the stage of National Rollout there was a reorganisation
and downsizing of the Horizon Team. I was retained by POCL in the autumn of
1999 to work on the second release of the Horizon system, ‘NR2+ (renamed
‘CSR+’) which provided the ‘Logistics Feeder Service’. I worked with the POCL
project manager on the development and testing of this release and was no
longer involved in planning and monitoring the National Rollout and continuing
live operation of Horizon.
17. By late January 2000 the contractual acceptance process was largely complete.
There were 9 medium and 48 low severity incidents outstanding and a ‘New
13 I 20
WITNO6090100
WITN06090100
Acceptance Process’ was documented and put in place to deal with these
(WINTO609_01/17). The Process transferred responsibility for the resolution of
many of those incidents to the development and testing teams and POCL
managers who with the completion of the NR2 (or CSR) release were now
working under the NR2+ (or CSR+) project.
18.CSR+ was a smaller and more manageable project that did not face the
complex issues described above with the first major Horizon release. Its
development and testing broadly ran to plan during 2000. Among my
responsibilities was to agree and authorise such documents as the ‘Service
Introduction for the Logistics Feeder System’ (LFS) (WINT0O609_01/18) which
set out how the deployment of the LFS system would be integrated with the
ongoing process of National Rollout of NR2. I was released from the
Programme in September before CSR+ went live.
Concluding Remarks
19. Finally, I will make some concluding remarks on my understanding of the
Horizon system at the time I ceased to be involved with the National Rollout,
and some further reflection I hope may be of assistance to the Chair.
20. Would it be correct to say that the Horizon system was rushed into live operation
without adequate testing and that this simple explanation accounts for the
problems that followed? From my personal experience, and the documentation
I have referred to, I do not believe this view is by any means accurate or
sufficient.
age 14] 20
WITNO6090100
WITNO06090100
21. During the period from April 1998 to late 1999 when I was continually involved
in preparing, revising and agreeing plans with ICL Pathway and the BA, I never
felt put under pressure by Post Office management to propose plans that I
considered high risk and driven by business or financial pressures other than
to achieve a successful delivery. The Programme plan issued in September
1998 sought to give adequate time to each major phase and included
contingency for potentially longer testing and trial periods. The Treasury Review
largely supported the plan and even suggested it might be possible to
accelerate it. As explained above, when the first cycles of testing in the autumn
of 1998 showed the system was failing on key requirements, an extended
period of testing was introduced and later the Live Trial was extended and the
Rollout delayed in order to give more time to resolve problems.
22.1 do not believe the situation, when I finished my work on the NR2 release in
autumn 1999, can be characterised as one where the Horizon system was
moving into National Rollout with lots of known serious bugs. The ‘hot list’
reports produced during acceptance did reveal a number of continuing
problems but there were no high priority incidents outstanding at final
acceptance. I am not suggesting, of course, that there were no faults in the
system. No major IT project can claim to deliver a 100% fault free system and
it would be naive to think that it could. However, the point I would draw attention
to in my statement is not that the Horizon project had to fix a lot of errors in
testing (that is what testing is for) but rather that the testing proved such a
lengthy and difficult process. I would like to explain this some more.
15 I 20
WITNO6090100
WITN06090100
23.As I have summarised above, the Horizon system with its key accounting
aspects was tested from August 1998 through to the end of the Live Trial a
whole year later in August 1999. Yet problems were still being found with such
key outputs as the cash account. The issue was not merely a matter of
individual errors but rather the underlying complexity and fragility of the branch
accounting system. This can be seen from examples in documents I have
already referred to. Here is a statement from my memo of 20" November 1998
(WITNO609_01/5):
‘errors are still outstanding on a significant number of the counter reports and
client summaries which are preventing effective internal reconciliation of data
within the outlet’ (emphasis in the original).
24.Here is another example from the memo I wrote two weeks. later
(WITNO609_01/6) that there was:
‘an action for POCL to define and agree additional business rules in the event
that mapping problems give rise to imbalances in the cash account (which
should only be an exceptional circumstance in live running).’
25. These point to endemic rather than individual faults in the system, that is they
arise from design issues not isolated programming flaws.
26. Similar points can be drawn from the Evaluation Reports for End to End Testing
and Model Office. Here is a quote from page 7 of the End to End report
(WITNO609_01/7):
16 I 20
WITNO6090100
WITN06090100
‘It was necessary to apply fixes to the counter environment to either move
forward, or to improve the quality of the test results ie. the incorrect cash
account mapping for a stock item would have caused mis-balancing Cash
Accounts in all offices.’
27. The practical implications of the complexity can be seen from this example on
page 7 in the Model Office report (WINT0609_01/8):
‘Although no fault in system functionality, certain areas of the solution (e.g.
Shared Stock Unit operation) proved difficult for live users to comprehend.’
28.A comparison I can draw with my wider IT axosiienes may help here. I
mentioned in my professional background that I worked on various accounting
systems at Coopers & Lybrand. When I think back on these systems they were
designed with a clarity of structure and content in their financial reporting, and
their underlying information flows. This had several benefits. It made testing the
system relatively straightforward and hence reduced the time to do so. Efficient
and effective testing made the live system more robust. If a fault was somehow
missed in testing, and occurred in live running, the client was also able to
identify it quickly and call for support. These were not characteristics that I
witnessed with the Horizon system.
29.While I was not personally involved in the specification or testing of the Horizon
system, the evidence I saw of the protracted difficulties in testing showed a lack
of transparency in how the system worked. This explains why the branch staff
had difficulties with cash accounts, and stock unit balancing, as reported during
the Live Trial. I have read the published testimony of some of the postmasters
17 I 20
WITNO6090100
WITN06090100
and postmistresses and I noticed that these raised the same issue. It was very
difficult for them to identify if, where and how errors had occurred in the system.
30.At the time I left Horizon I was not so much worried as to whether known faults
had been fixed. It was rather that the system delivered into Rollout had an
ongoing vulnerability to error due to its complexity and lack of transparency.
When errors arose over time during live operation, as they do in all systems, it
would be difficult if not impossible for postmasters and postmistresses to
understand what had gone wrong. So a situation arose where they could not
validate the integrity of their own financial information, and yet they were held
accountable for it:
3
=
.I trust this statement is helpful to the Inquiry and I would be pleased to offer
whatever further help I can.
Statement of Truth
I believe the content of this statement to be true.
GR
Signed:
Dated: mh Sepkember 2022
age 18I 20
WITNO6090100
WITNO06090100
Index to First Witness Statement of Andrew John Dennis Simpkins
No. I Exhibit Number Document Description URN
1 WITNO609_01/1 Horizon Level 1 plans POL00028437
2 WITNO609_01/2 Responses to Concerns I POL00028436
briefing note included with
other documents — under
‘Letter from Andrew
Simpkins to Vince Gaskell’
but not sent with the letter
3 WINT0609_01/3 Presentation to CAPs Board I POL00028411
— 15" January 1999
4 WITNO609_01/4 Proving the TIP Interface I POL00028435
Draft Report
5 WITNO609_01/5 Briefing Note on Status of I POL00028431
Testing - 20 November
1998
6 WITNO609_01/6 Horizon Testing and I POL00028429
Programme Plan - Current
Status 4" December 1998
7 WINTO609_01/7 End to End Testing I POL00028419
Evaluation Report
8 WINT0609_01/8 Model Office Testing I FU00058445
Evaluation Report
9 WINT0609_01/9 Letter from Stuart Sweetman I POL00028405
to Peter Mathison
10 I WIN9TO609_01/10 I Horizon Management Team I POL00028372
Meeting — Managing the Live
Trial
11 I WINTO609_01/11 I Acceptance Incident Hot List I POL0028355
12. I WINTO609_01/12_ I Management Resolution I POL00028332
Meeting — 12" August 1999
13. I WINTO609_01/13 I Acceptance - Additional I POL00028344
Offices ees
I WINTO609_01/14 I Acceptance Update 4 —I POL00028473
I Output and Actions
15 I WINTO609_01/15 I 1999 Rollout POL00028461
16 I WINTO609_01/16 I Extract from — Rollout-out I POLO00028559
Decision paper in the papers
for the Roll-out 2000 Pre-
ze 19 I: 20
WITNO6090100
WITNO06090100
Checkpoint Meeting of 15"
November 1999
17. I WINTO609_01/17 I The = New — Acceptance I POL00028508
Process
18. I WINTO609_01/18 I Service Introduction of the I FUJO0001385
Logistics Feeder System
20 I 20