Often maintenance systems don't reap the benefits that they promise 
through no fault of their own.  How can you expect a system to improve 
underlying data?  The answer is that you can't.  What you need is to 
have good data in the system so that it can be accessed, processed and 
used to provide practical information for the organization.
Let
 me illustrate the cost of not having good data with an example.  A 
multi-site manufacturer has four locations, three of which are in fairly
 close proximity to each other.   Each site has its own autonomous 
storeroom with inventory parts.  At each site, there is a part time 
catalog manager responsible for all database activity.  Because the 
plant is unionized and positions often change, the catalog manager may 
be replaced every few months.
The resulting inventory catalogs 
reflect this: inconsistent manufacturer naming; missing manufacturer 
part numbers; inconsistent use of symbols/abbreviations; spelling 
mistakes; incomplete descriptions and; duplicate items.  System word 
searches are next to impossible and finding a part is a frustrating, 
challenging, usually unsuccessful experience.
Maintenance workers 
at all locations had long lost faith in stores; each kept a stash of 
parts hidden somewhere for his own use.  To plan for a repair job, they 
would attempt to find parts through the system, but if unable to locate 
what they needed, they would abandon the search and just order the part 
directly; in the case of an emergency, they might call another location 
to request the loan of a part.   Inventory value across the company 
topped $80 million.
Recognizing that something had to be done, the
 company attempted to undertake the data cleaning themselves.  They 
established a team of nineteen internal people comprised of maintenance 
workers (Electrical, Mechanical, Instrumentation & Pipe Fitters) 
from all four sites as well as two support people and one Inventory 
Specialist.
After more than a year of effort, and with only half 
the database cleaned, they decided to engage outside data cleaning 
experts to revitalize the effort.  Systematically, the data from each 
site was cleaned.  In conjunction with maintenance workers from all 
sites, a common layout for item descriptions with acceptable 
noun/modifier pairs was developed; the order of attributes was 
negotiated to satisfy all locations; terminology, symbols, abbreviations
 and industry nomenclature were agreed upon.  It took six months to 
rework the entire database.
Having good data brings with it 
measurable rewards.  Duplicates within sites were revealed to be in the 
10% range.  Common items across sites were identified in the 25% range. 
 Merging the three regional stores into a central warehouse reduced 
overall stocking levels and allowed sites to share common critical 
spares.  It also freed up millions in cash savings.
Item searches 
successfully revealed part information that maintenance workers could 
count on.  As confidence in the central stores grew, additional stock 
from private caches was repatriated, further adding to the savings 
realized.  Overall, across the company, inventory was reduced by more 
than 20%.
The data cleansing effort clearly paid for itself 
several times over.   It also became the impetus for other corporate 
initiatives.  The company went on to improve its item-equipment links to
 further enhance the maintenance system.  In addition, it consolidated 
items along product lines and reduced its supplier base for volume 
discounts.
