HCi Journal of Information Development
By Onno van Ewyk
This is the true story of a transformation that took place in the manufacturing operation of a multi-national pharmaceutical company. This transformation took it from being an efficient, well-run, and people-centric operation to an even more efficient and better-run people-centric operation.
It is a story with happy beginnings but a sad ending. There are two heroes - a recently appointed young manufacturing manager and his assistant - and three villains - the QA manager, an accountant, and a multi-national strategist.
The story begins when the manufacturing manager and his assistant decide to radically upgrade the way they manage the operation. Although the plant was already running efficiently by world standards using traditional conservative bureaucratic methods (including a benevolent and supportive attitude to staff), it was under threat of closure because of a world-wide rationalisation of manufacturing operations that was taking place.
The manager and his assistant felt that the best way for them to respond to this challenge was to make their plant not only good, but the best in the world. They planned to do this by introducing a continuous improvement approach that bolstered teamwork and encouraged creativity moderated by disciplined measurement and analysis. The aim was not to overturn or replace what they were currently doing, which was working well enough, but to boost it and if possible raise performance to a higher level.
The manager engaged a consultant to help design a program for introducing the new approach. They decided to group the staff into ‘natural work teams’, that is, those who performed particular functions together, and to appoint team leaders from the supervisor-level staff. This meant that the new approach would align, and therefore not conflict, with existing authority and organisational reporting lines. Time was set aside during regular working hours for the teams to meet and discuss how their teams were performing and how they would make improvements.
Having set up this operating framework, a training program was designed and implemented to give the staff the necessary skills. All 50 staff members attended two days of training, which covered small-group skills (teamworking), analytical methods (customer supplier mapping, flowcharting, pareto analysis, cause and effect diagrams, statistical process control) and creativity tools (nominal group technique, brainstorming).
The manager decided that the program should have a name, so it was called SCIP. The S stood for the company name, let’s call it Summer Pharceuticals, and the CIP for Continuous Improvement Program.
From the outset, the QA manager was involved in the program. The manufacturing manager and the consultant argued that quality assurance could only benefit from the program because it would focus all staff in a direct and practical way on quality issues. Outwardly the QA manager agreed, but it subsequently became apparent that his support was not whole-hearted.
After the training, the teams began their work in earnest.
They met, talked, analysed, agreed, and acted. Not everyone displayed the
same level of keen-ness and commitment. Each
SCIP team comprised 6-7 people. Of these, typically, there was:
1
disbeliever – this person was unwilling to help the company, “I just
want to do my job, management is paid to do the thinking”, was their
attitude
1-2
passengers – these people were happy to take the time off work, but
contributed little to the discussions
1-2
contributors - these people participated positively during the team
sessions, but did not take many direct initiatives
1-2
enthusiasts - these people pushed, came up with ideas, and worked hard to
see them implemented
Even with this level of participation, which could be expected to improve as team leaders grew more skilled over time, the impact of the SCIP teams was immediate and impressive. It is worth looking in detail at some of the things they achieved during the first year.
The
output of the tablet compression machine had always been collected in 18
Kg plastic buckets. These were heavy and awkward to handle, particularly
for the women who comprised most of the workforce in the area. These
buckets could not be lifted up to the hopper of the packaging machine.
This hopper was so high off the ground it required a step ladder to reach
it. As a result, tablets had to be transferred using a scoop. This was
slow and often caused damage to the tablets, which were easy to chip or
break.
Another
problem was that the bucket size had naturally become the de-facto unit of
measurement for production runs. This meant that smaller packaging runs
were often wrongly estimated, requiring already packaged tablets to be
unpacked and repacked in other boxes to satisfy orders.
The
SCIP team involved with this process examined the problems and proposed
that 10 Kg buckets be used instead. They put their arguments to
management, who agreed and gave their approval to replace the buckets.
This cost less than a thousand dollars.
The
packaging operation was speeded up (no precise measurements were taken
because the advantages were so obvious), damage to tablets dropped to
virtually nil, there was now a significantly lower risk of injury to
workers handling the buckets, and smaller packaging orders were now
consistently accurately filled.
There
was an unexpected benefit after the smaller buckets began to be used. Each
bucket was normally weighed after filling from the compression machine. A
metal detector on the compression machine required a minimum drop
distance. The smaller bucket allowed the weighing machine or balance to be
placed under the bucket. Previously the larger bucket had to be
transferred to the balance after filling. With the smaller bucket being
filled on the balance, handling was reduced and workers were able to fill
each bucket with a consistent weight of tablets.
The
plant had a new dispensary for ingredients that was fitted with high
volume dust extraction in order the keep the air clean. The system was so
efficient it was extracting 500 Kg of raw material each year, worth
$10,000. The SCIP team achieved a 60-70% reduction in this waste by
designing and fitting a collar to the raw material bins to change the air
flow, and by experimenting with the extraction settings to achieve the
required air cleanliness while minimising raw material losses. The cost of
making the collars was approximately $100.
300,000
packs of ampules were received in bulk form each year. These packs were
then packed into boxes by hand. It occurred to the SCIP team that the
blister line's high-speed packaging machine could be used to pack the
ampules automatically. It was previously not thought possible by
management to use this machine for ampules. After some discussion and
experimentation by the team, they decided to fit a new detector. This was
done using in-house designed and manufactured tooling . The cost was
approximately $2000.
The
changes made were successful. By automating this previously manual
packaging task, 2000 hours of direct labour time was saved, which
translated into $40-50000 dollars in cost (casual labour was used, so no
full-time jobs were endangered by the change). Throughput was six days
quicker.
Two
different machines were traditionally used to count tablets and capsules.
The 'tablet' counter was much quicker than the 'capsule' counter. The
'capsule' counter was brought out whenever the capsules were to be
packaged. This was relatively infrequent - 4 times per year, requiring 9
days on the packaging line per occasion at the rate of 130 packs per hour.
In
factory parlance the two machines were known as "the tablet
counter" and "the capsule counter" but they were
essentially the same in function. With the introduction of SCIP, current
practice was no longer taken for granted. A new franchise for change
opened up the possibility that both machines may be thought of as
"counters" and the most efficient one should be used as much as
possible. (SCIP not only affected mindsets, but mobilised changes which
had previously been mooted but were never followed up.)
For a
cost of $8000, a new set of slatts was purchased for the newer machine to
allow it to handle capsules. This increased packaging speed for capsules
from 130 per hour to 2000 per hour. The saving in direct labour costs was
$12000 per packaging run. This was a significant saving, but the principal
benefit was that packaging run times were reduced from 9 days to one. This
made the packaging line free for an additional 32 days of production per
year. The quicker turnaround also meant that tablet orders no longer had a
potential additional 9 day wait (when capsules were being produced), so
tablet inventory could be reduced with affecting order fulfillment.
There
was a high level of variability in tablet coating times. For average 10
day runs, the time taken could be plus or minus 3 or 4 days. This meant
scheduling other processes dependent on coating was very difficult. Tablet
coating runs took place only 4 times a year but they were very disruptive
to packaging schedules.
Efforts
were made to trace the sources of variation but these proved elusive.
Times could not be traced back to either ingredient variations or machine
settings. The variation could not be reduced.
The
only benefit resulting from the exercise was that they found that
scheduling was more accurate if modal rather then average coating times
were used in production time estimates.
RF
(Radio Frequency) terminals for use by warehouse picking staff had been
considered for some time. The warehouse staff decided to use SCIP tools to
implement the project and formed a special team. It took two months to
implement the new system. Going on previous experience, it would have
taken 6 or 7 months without the use of the SCIP team approach.
This
project was an example of process re-engineering. Prior to the RF terminal
introduction, orders were entered into the computer system to produce a
picking slip. This was placed in a vacuum tube and sent to the warehouse.
The warehouse operator would then take the picking slip into the warehouse
and pick the goods listed, ticking off each order line as it is picked,
and noting any "out of stock" adjustments on the slip. The slip
was then returned for invoice generation. The picking slip was easy to
misread, and pharmaceutical goods are notoriously difficult to pick
because products often only differ in terms of pack sizes and
concentrations. Names of products also tend to be very similar.
Any
adjustments slowed invoice printing, so in many cases goods had to be
shipped without invoices, delaying payment times and incurring additional
mailing, handling, and reconciliation costs.
With
the RF terminals there is no delay between order entry and despatch of
information to the warehouse. Warehouse operators are able to pick goods
line by line from the terminal display and confirm that the correct goods
have been picked using a bar code reader. Any "out of stock"
adjustments are entered there-and-then by the operator, adjustments are
made instantaneously by the computer and invoices printed immediately. As
a result invoices are always sent with the goods.
Prior
to the RF terminals, picking errors were 3-5% in the worst months. These
errors were reduced to an average of 0.1% per month, and these were
restricted to quantity errors only (incorrect goods errors were completely
eradicated).
A
proposal was put to management to re-arrange the packaging room and
refurbish it by installing different feed mechanisms. The cost of this was
estimated at $80,000 and the projected savings were estimated at $50,000
per year (half a person and increased throughput times).
Head
office approval was required for all capital expenditure over $10,000.
Approval could not be secured, not because the proposal was unsound, but
because there was a blanket ban on capital expenditure.
In overall terms the results of the program were spectacular. Easily measured cost savings on initiatives taken in the first six months totalled $70,000 per year on total outlays of only $10,000. Further savings of $50,000 per year were identified but required capital expenditure to realise them. In addition, very significant other benefits were secured in terms of difficult-to-measure savings, quality improvements, and reduced risk of accidents or damage. Consulting and training cost $24,000.
Benefits could be expected to increase significantly as the program matured. Most of the initiatives were reported as coming from the combination of new teamwork skills, observation, and being given a franchise to change things. The hard-edged analytical tools, including statistical process control, were yet to make an impact.
SCIP achieved its success despite some internal resistance. The QA manager, although outwardly cooperative towards the manufacturing manager and the consultant had, it appeared later, always been anti the program. His view was that anything to do with quality was his preserve and that the idea of engaging staff to help improve processes was "a wank". Despite his antagonism, his staff, who had all been through the training, were enthusiastic. They were assigned to various teams and wanted to continue to work with them. The QA manager resisted at every turn; "I do not want to see anyone doing anything that there is not an SOP for", he was often heard to say.
The
warehouse manager left the company a few months after program
implementation. He was replaced by, "an accountant with a chip on her
shoulder", as she was described by one staff member. She was
interested only in cost-cutting and analysing the numbers. She had no time
for teamwork, SCIP, or its aims, which ironically was her best strategy
for achieving what she was looking to do.
But
these negative forces did not bring about the end of SCIP. It had built up
enough momentum to carry on through this resistance and would probably
eventually have overcome it through its demonstrative success and
enthusiastic champions. The end came when a multi-national strategist
decided that, at 0.2% of international production capacity, Summer
Pharmaceuticals could afford to do without its Australian plant so it was
shut down at a de-commissioning cost of $11M.
First published 2003
This article may be reproduced only with the permission of HCi . Copyright HCi, 2001-3.
HCi information development - www.hci.com.au
(technical writing, quality management, knowledge management)