POL00029991
POL00029991
IN THE HIGH COURT OF JUSTICE CLAIM NO. HQ16X01238
QUEEN’S BENCH DIVISION
BETWEEN:
ALAN BATES & OTHERS
Claimants
-and-
POST OFFICE LIMITED
Defendant
WITNESS STATEMENT OF RICHARD ROLL
I, Richard Roll, of} GRO 'WILL STATE:-
1.
Between 2001 and 2004 I worked for Fujitsu Services as an IT Product Specialist,
focussing almost exclusively on the investigation and resolution of issues with the Horizon
system. I was based at a Fujitsu office building in Bracknell, Berkshire.
My particular role at Fujitsu was 3°4/4" line support and development on Post Office’s
Horizon system. This involved me investigating and resolving a range of IT system issues
and queries, which hadn't been resolved at the 1*t or 2"4 support lines.
I recall that the Post Office IT system serviced by Fujitsu involved around 40,000 users at
around 20,000 sites throughout the UK. This was known as “the Estate”
As is well known, the Horizon system is a substantial IT system. The system has had
numerous versions “layered” onto it, with extensive and ongoing tailoring and
modifications carried out by Fujitsu IT professionals. There were many Fujitsu IT
personnel based in the same building as myself in Bracknell.
I recall there being two “server farms”, which mirrored each other, and which held data
from all UK Post Office branches serviced by the system. There were also security
servers, network control systems and data communications systems.
My role involved receiving issues or problems being experienced by people working within
Post Office branches. In many cases, this involved me dealing directly with sub post-
masters and others employed in Post Office branches.
By way of example of the type of issue that I would deal with, if a financial discrepancy
had arisen in a branch (e.g. a “shortfall” of oa. chdt han I woul ABD iherk sequentially
~ GRO \
10.
11.
12.
13.
POL00029991
POL00029991
through all transactions over the relevant period, and also work through thousands of lines
of computer coding. Software programs were written by us to strip-out irrelevant data, to
enable us to more easily locate the error.
lf there was a single error then that would be easy to identify, however there were often
multiple errors which would “snowball”. This would make matters more complicated. Any
errors made by sub-postmasters would be relatively easy to identify, and would normally
be picked up by 1 or 2 line support. If an error was referred to us then it was extremely
unlikely to be due to a mistake made by a postmaster; the vast majority of errors I dealt
with were due to coding errors or data corruption.
We regularly identified issues with the computer coding in the Horizon system. We would
then flag those issues to the Fujitsu IT software developers. The developers would then
work on a “fix” while we monitored whole estate in relation to that issue.
My recollection is that the software issues we were routinely encountering could, and did,
cause financial discrepancies at branch level, including “shortfalls” being incorrectly
shown on the Horizon system. !f we were unable to find the cause of the discrepancy
then this was reported up the chain and it was assumed that the postmaster was to
blame.
I would estimate that there were over 30 individuals working on the same floor as me
(floor 6) in the Fujitsu building at Bracknell. My recollection is that many of those
individuals were involved in similar work to myself, or were involved in other Horizon
related IT work. I would describe much of the work being carried out as “fire fighting”
coding problems in the Horizon system.
There was a perception, certainly in the locality of where I was working, that the Service
Level Agreements between Post Office Ltd and Fujitsu involved financial penalties
payable by Fujitsu to Post Office. A percentage, I think 99.5%, of all transactions carried
out in the Estate had to be processed within 3 working days (for example, if someone paid
a telephone bill, the money had to have been debited from their account and transferred
to the telephone company’s account within 72 hours). Additionally, any discrepancy in the
post office accounts had to be resolved speedily — I do not recall the exact timeframe.
The Fujitsu software developers would usually manage to produce “fixes” to the coding
issues that arose. However, due to multiple upgrades and data volume, there were only a
limited number of time windows when software updates could be released into the Estate.
For example, there could be six weeks delay before a fix could go out, and during that
period Post Office branches could continue to be affected by the coding issue. I also recall
situations where software developers worked on a coding fix which was then sent out,
however the bug reappeared several weeks later because, it seemed, the IT team
responsible for developing upgrades had been working on an older version of the software
which had not received the coding fix:
POL00029991
POL00029991
14. As well as software issues, I can also recall that there were regular IT hardware issues at
branch level. However, I would reiterate that the main recurring issues were software
issues.
15. During the course of resolving the software issues, we would frequently access a Post
Office counter IT system remotely. An example of a relatively common problem that arose
was when a binary bit would “flip”, thus a “1” became a “0”. By way of example, I would
contact the branch and arrange for them to stop using the computer for a limited period of
time, then I would log into the branch’s system, download all the data from the relevant
computer, identify and correct the corrupted line of data, then log back in to the computer
at the post office and transfer the corrected data to the branch computer to re-create the
database, thus correcting the problem. Correcting the data could be done in a matter of
minutes, or it may have taken hours, depending on the complexity of the problem. The
act of transferring data on to the post office computer (that is, the time from ‘pressing the
button’ so to speak, to completing the data transfer) actually took only seconds.
16. Still on the subject of remote access to branch systems, as I recall some errors were
corrected remotely without the sub-postmaster being aware.
17. Also, sometimes we identified an error that the postmaster was unaware of, we would
then log in to his system while it was switched on but not in use, download and correct the
data and prepare it for uploading back on to the post office computer, then call the
postmaster to inform him that there was problem and that we needed two or three minutes
to correct it. If the postmaster started using the system while we were downloading the
data to our computers then he would not have been aware that we were logged in, but it
would interfere with our download and we would have to start again later.
18. The ability to remotely access the Horizon system at branch level was extensive, in that
we were able to change not only data and transaction information, but we also had the
ability to insert transactions and transfer money remotely without the sub-postmaster
knowing. Obviously this was not done by me, however I can recall thinking that a third
party may have been able to do that if they could have remotely accessed the system in
the way that we could (which may or may not have been possible).
19. In summary, the issues with coding in the Horizon system were extensive. Furthermore,
the coding issues impacted on transaction data and caused financial discrepancies on the
Horizon system at Branch level. It was those issues that I, and other colleagues at Fujitsu,
were routinely working on daily. Furthermore, remote access to the Horizon system at
Branch level was extensive, as was the ability to, change data and change transaction
. being aware
of this.
Statement of Truth
POL00029991
POL00029991
MrRichardRolls—i—i‘“‘sé™SOSCS*~*~<‘“‘;‘;*~*”
Dated the / of Sly 2016