and re-motivating a demoralised IT unit. The remainder of this chapter
describes how this was done and the lessons learned from this experience.
2 BACKGROUND AND CONTEXT
Agribusiness, food and fi
fisheries is Ireland’s largest indigenous industry
employing some 150,000 people with an annual output value of over €24
billion. Nearly 90 percent of Ireland’s beef and dairy produce is exported to
markets worldwide and many of the operations of the industry depend on
real-time access to the ICT systems of the Irish Department of Agriculture,
Fisheries and Food.1
The DAFF operates in a particularly challenging external environment
due to the nature of the Common Agriculture Policy support schemes and
the ever-present risk of animal health and food safety crises. CAP payment
schemes are complex and have demanding implementation timelines with
A Green Revolution 49
which all member states struggle to comply. Any shortcomings in terms of
controls or timeliness can incur EU-imposed fi
financial penalties that can
potentially run to hundreds of millions of euro. Animal health crises can-
not be predicted and they must be responded to without delay. Food safety
incidents must likewise be dealt with decisively to protect human health
and the reputation of Irish food and valuable export markets.
In addition to the above, the DAFF is responsible for smooth operation
and delivery of approximately €3 billion per annum in fi n
fi ancial supports
to farmers and agribusinesses as well as for animal disease testing and the
tracing of farm animal movement from birth through to entry to the food
chain (or death). In Ireland there are two million bovine births and seven
million bovine animal movements each year. In addition to all of the other
challenges faced in 1999, the department was going to have to change its
ICT architecture and essentially all of it business applications without any
disruption to its on-going operations.
To further complicate matters, many people and organizations act as
trusted agents of the department. These include private veterinary practices,
livestock marts, meat/slaughter plants, major food processors and export-
ers, providers of farm management software packages, vendors of animal
tags, outsourced animal movement bureau, forestry companies, farm plan-
ning providers that submit claims on behalf of farmers and the Irish Cattle
Breeders’ Federation. These groups need access to specific information. For
example, approximately 15 million on-farm disease tests are carried out
annually by approximately 600 private veterinary practices (PVPs) across
Ireland. The results of these tests must be recorded against animal records
in the department’s animal health database. Each proposed movement or
slaughtering for human consumption is then pre-checked against this data-
base to determine if it is permissible. The extensive interaction between
PVPs and the department was highly labor-intensive. To add to the prob-
lems, the system was based on dated county-by-county mini-computers and
recorded exceptions only rather than all test results.
3 A
PROGRAM OF PARALLEL STREAMS
Turning around ICT in the above circumstances required simultaneous
action on several fronts. This chapter describes four important aspects of
this process concentrating primarily on the years 2000–2004. These key
fronts were technical, political, operational, and organizational, each of
which is now discussed.
3.1 Technical Strategy
On the technical front there were several sub problems that needed to be
addressed in parallel. These are illustrated in Table 5.1.
50 Frank Bannister, Regina Connolly, and Philip O’Reilly
Table 5.1 Overview of Strategy
From the outset, a number of principles were agreed that underpinned all
subsequent decision making about the technical aspects of the plan:
1.
Existing operations could not be disrupted, even though temporary dis-
ruptions would have made implementation easier at several points.
2. The architecture was envisaged in a number of layers (Table 5.2). It was recognised that the data layer would be the most critical part of the new
architecture as anything else could be fi
fixed in retrospect much more
easily than problems in the data structure, and it was essential that data
could be re-used where it occurred across the various lines of business.
Table 5.2 Multilayer Architecture Used
A Green Revolution 51
3.
Consequently there would be one corporate data model and within
this three key entities would be: customer,
r animal, and land. The leg-
acy corporate data structure inherited at the start of the project was
highly fragmented. Clearly, collecting data once, getting it right and
re-using it as needed would yield major effi
fficiencies and improve data
integrity. It was therefore decided to develop a high-level corporate
data model and let whichever application project fi
first needed a key
entity take responsibility to fully fl
flesh out that aspect of the model
while taking account of future corporate uses.
4. While legacy systems would be maintained for the time being, the
department would move to emerging technologies for its new applica-
tions with a view to doing the same for legacy systems as each one
needed a major re-write. Emerging technologies were monitored, and
opportunities for each technology to deliver operational benefits were
examined or trialed.
5. A policy known as “deep e-government,” linking back offi
ce
ffi systems
directly to customer systems where possible, would be adopted. This
was in contrast to the front-end orientation of most e-government at
the time.
6.
The
fi
five layers of the architecture would be addressed on a phased
basis. The top layer, that is the business process architecture, was
not to be attempted at the outset and would be revisited. The overall
architecture of the business applications, the second layer from the
top, was only addressed to the extent that there would be a corporate
customer system as a hub for all applications, and that there would
be a single view of the customer, a single view of the land and a single
view of the animal. It had been determined that the data layer would
be key to the future and would be addressed fi
first.
7. The technical environment would be procured on the basis of bundled
solution incorporating the underlying infrastructure, operating sys-
tem, and development environment. The department did not have the
resources to dedicate to selecting and integrating best-of-breed compo-
nents and would have found such an approach diffi
cul
ffi
t to support.
8. All of the corporate business applications should be built using a
three-tiered approach (user interface, application,
and data) with a
standard desktop browser providing access for internal users. Consis-
tent style sheets would be developed for the user interfaces.
Beyond the above, the overall architecture of the complete portfolio of
business applications was not specifi
fied in detail at this stage.
3.2 Political Strategy
Given the history of ICT in the department, achieving and improving inter-
nal morale in the ICT group and maintaining external credibility were
52 Frank Bannister, Regina Connolly, and Philip O’Reilly
going to be critical success factors from the outset. One way to deliver a
quick win was on the desktop. This was done by installing a standard well-
specified PC with a set of offi
ffice products and a standard browser to front
the corporate applications. Whereas there were as yet no new corporate
systems, the PCs could be used to access legacy applications. More impor-
tantly, staff n
ff o longer had problems with e-mail or incompatible versions of
offi
ffice tools and the new LANs and WAN meant no interruptions in service.
Coupled with the DAFF Intranet— eZone, and Internet access, all staff now
had easy access to up-to-date information. The psychological impact of
the new machines and the local tools were signifi
ficant. ICT was delivering
something that staff could see and use.
Other quick wins included Short Message Service (SMS) from the desk-
top and videoconferencing. As events unfolded, having these desktops
and offi
ffice servers, a newly architected network (WAN and LANs) and the
SMS facility in place proved to be fortunate indeed when Foot and Mouth
Disease arrived in Ireland in March 2001 (see Section 4). These successes and others built further confi
fidence throughout the organization in ICTs
judgment and ability to deliver working solutions for the business needs of
DAFF reliably and on time.
3.3 OPERATIONAL STRATEGY
Conventional 1980s structured methodologies assume that the ICT applica-
tion developers gather requirements from a line of business unit and trans-
late those into detailed functional specifi
fications that are signed off
ff and then
built (Gasson, 1995). The view taken in DAFF during this program was
that conventional structured methods would be too slow and labor-inten-
sive. Furthermore, with such an approach there is a considerable likelihood
of looping-back at testing stage to correct misunderstood requirements
leading to cost and time over-runs and sub-optimal application structures
(Morris, 1990). Instead, teams were established which combined business
specialization as well as ICT skills. Contrary to the prevailing view (see
Ross and Weill, 2002), the lead ICT person was expected to understand
fully current business requirements as well as possible future scenarios. By
understanding the business, the ICT lead person could anticipate where
fl
flexibility would be likely to be required in the application and in particular
the implications for the underpinning data structures that would have to
support the business and application change into the future.
A good example of this approach in action was the Corporate Cus-
tomer System. The 2000 strategy had included a proposal that all opera-
tional systems would share common customer data. In order to achieve
this there would need to be a single system or set of functionalities to
create records of all customers and all of their relationships. This should
include a rich view of all the customer relationships with the organization
A Green Revolution 53
and in the case of DAFF would extend to the relationships with herds or
other entities with which the department dealt.
A decision was therefore taken early in the project to introduce a strate-
gic program to deliver this type of functionality. Since this program would
be long-term and the system would be central to the overall group of opera-
tional systems, it was decided that the project management, design, and
build leadership would be carried out by in-house ICT staff with contract-
ing restricted to the programming level. The outcome was a system which
rapidly became a core component of the department’s operations and which
remains so to this day.
The deep e-government principle was applied to all major developments,
streamlining the operational processes. During the development of the ani-
mal health computer system this principle eliminated the vast volumes of
paper fl o
fl wing between the DAFF and private veterinarians.
3.4 Organizational Strategy
An immediate problem that had to be faced was the shortage of inter-
nal resources. In 2001 there were two particularly large-scale application
projects that needed to get underway (integrated mapping and payments
system (iMAP) on the CAP area aid based payments and the animal
health computer system (AHCS)) and there simply were not the skilled
in-house resources to do these, so the outsourcing approach was the only
feasible option.
For these initial projects, to try to ensure knowledge transfer and post-
implementation control, in-house resources were deployed as members of
the team. For subsequent projects, in-house resources were assigned to
project management and technical lead positions on all high value tasks of
strategic long-term importance.
In assigning ICT staff t
ff o various roles throughout the program, consid-
erable attention was paid not just to their technical skills and experience,
but also to their personality and work-style preferences. This approach in
assignments resulted in a more productive workforce.
4 PAYBACK
In the period between 2000 and 2004 alone, several major crises/develop-
ments were to test the eff
ffectiveness of the new computer systems and the
department’s ICT capability. Four of these will be briefl y con
fl
sidered.
By far the most signifi
ficant, in national terms, was the outbreak in March
2001 of foot and mouth disease (FMD) that spread quickly throughout the
United Kingdom. Ireland as a country has FMD clear status. Any loss of this
would have had catastrophic consequences for the economy. FMD is air-
borne and highly contagious. With so much movement between the United
54 Frank Bannister, Regina Connolly, and Philip O’Reilly
Kingdom and Ireland, an outbreak was inevitable, and this occurred on one
farm on the northeast coast of Ireland. Reaction from the department was
swift. Movements to and from the infected areas were traced and all herds
within a chosen infection zone were visited, tested and culled where neces-
sary. This enormous operation required the setting up of a Local Disease
Control Center with access to all of the department’s ICT systems including
full customer information, herd profi
files, and spatial mapping information.
The successful confi
finement of FMD in Ireland reinforced the fact
that ICT
was central to the department’s operations, and, secondly, its own in-house
ICT department was capable of delivering under pressure.
A consequential development was a decision to tag sheep. Ireland had a
world-class traceability system for bovines that underpinned the reputation of
quality Irish beef. Sheep, on the other hand, were neither tagged nor traceable
at this time. In the immediate aftermath of FMD, it was decided by the then
minister that all eight million sheep in Ireland would be individually tagged
enabling full tracing of sheep movements within a 2-month period.
Involving just four people—two ICT and two line-of-business staff—the
business process and the design of the system was completed in 2 days,
specified and built in 4 weeks, and rolled out in a further 4 weeks. The
system, subsequently known as STAR (Sheep Tagging and Registration),
went live on time and managed the issuing of fi
five million tags in its fi
first
month of operation. In addition to consolidating the ICT division’s repu-
tation for delivery, it confi
firmed the value of deep e-government since the
process fully integrated the business processes of the tag vendors with the
DAFF systems.
A third example of the impact of the new regime was demonstrated
when the EU committed to a mid-term review of the 2000 CAP reforms
in 2004. The CAP reform off ere
ff
d member states a number of options and
Ireland chose an option that was the most economically attractive. The
government decided to implement this in 2005. Due to the fl
flexible struc-
tured data model developed for the Corporate Customer System and the
linkages that it allowed for recording all periods of ownership of holdings
over time, it was possible to accurately calculate payments, thus positioning
the Department as a leader among EU CAP paying agencies.
The fi
final major unanticipated development during this period was an
extension in the department’s remit. During the delivery of single farm pay-
ment system, a cabinet reshuffl
ffle resulted in the responsibility for Forestry
being transferred to the DAFF. The department from which it transferred
had planned on embarking on a major project to integrate the management
Public Sector Transformation Through E-Government Page 10