|
|
|
Cover Story
|
| |
|
|
The oil industry has long been recognised
as one of the leaders in the field of KM, with
companies spending millions of oil dollars on
promising initiatives. Campbell McCracken
looks at what is in the pipeline.
|
|
Sometimes
crude, sometimes
slick | |
Much
money has been spent by oil companies creating 'best
practice' databases. "A lot of these are what we call a
storehouse model," says Sopheon's CEO Andy Michuda,
"which means ‘let's go capture everything we know about
our processes, our policies, regulatory law, etc, and
put it in a database.’"
The
idea is that once the information is captured, people
can just go to the database to find out how to solve
particular problems instead of re-inventing the wheel
each time.
But
these solutions have their drawbacks. "What you find
constantly when you investigate," explains Michuda, "is
that these databases generally are not kept current, and
they're not actually engaged by the user community, as
they are a one-off and are hard to find."
Eric
Lesser, executive consultant at IBM agrees: "A
repository needs to be moderated in some way because
otherwise it becomes like a garden that is allowed to
grow wild. If it's not seeded properly no-one will look
at it. If it's got too much information in there no-one
can determine what kind of value it has."
Ageing demographic However, the oil industry is facing an even
greater challenge over the next five or ten years,
because of the demographic profile of its workforce.
"The problem is that most of these people are in their
mid-fifties," says Sopheon's Michuda. "The organisations
know they're facing a knowledge drain, so they're
getting companies like us to work with their team of
people to start capturing their know-how and their
understanding." (See
Box: Sopheon at GTS-Geotech.)
"A
lot of senior experienced people are going to be
retiring in the next ten years," says IBM's Lesser.
"Plus you also have fewer and fewer young people who are
interested in pursuing these careers, so the number of
petroleum engineers graduating is a tiny fraction of the
overall demand from the industry."
The
threat of losing experts has meant that companies have
to capture the knowledge before it goes out the door. At
the other end of the workforce chain, companies face
significant cost pressures to accelerate the training of
new recruits to make them become productive more
quickly. "When you look at all these demographic
factors, it's frankly quite frightening for some of the
oil companies." says Lesser.
Mergers and acquisitions Recent mergers within the oil industry, such as
those between Texaco and Chevron and between Exxon and
Mobil have also created problems. When companies merge,
how do employees know where to find information such as
best practices and company policies? One of the side
effects of a merger is to increase the number of
disparate database systems within a company, making it
very difficult for the IT people to be able to present a
unified view across all systems.
"The
biggest problem these guys have is deploying federated
search across the enterprise," says Hummingbird's VP of
US marketing Peter Auditore. Federated search includes
searching structured databases, such as online
transaction systems, data warehouses, in addition to
unstructured data. "One of the other biggest layers
under this is indexing and categorisation of all of your
stuff. That’s a big, big component."
"This problem is simply too large for basic
search engines to address satisfactorily," says Inxight
Software's sales manager Mike Tarttelin. Inxight's
automatic categorisation software works in conjunction
with a company-defined set of key concepts. These can be
arranged either as a full hierarchical taxonomy or as a
flat set of key concepts to define a classification
system for all information.
"The
Inxight stuff has built in intelligent agents that tell
you when your categorisation has extended beyond your
set nomenclature or taxonomy," continues Auditore. "If
something new comes up it will send you a message saying
'Hey you might want to create another category in
the index for this particular area that is new to the
system.'"
|
|
Box 1 Sopheon at
GTS-Geotech |
|
GTS-Geotech is an IT
services and consultancy company operating in four
continents, sometime is remote locations. With
almost its entire staff of 100 consultants
scattered around the world at any time, it was
difficult for someone in the field to find answers
when they came across problems.
"As in most
organisations, there are some senior people with
an awful lot of knowledge," says GTS-Geotech's
senior consultant Steve Kaye, "and there are some
lesser skilled individuals who are training up.
But they're still on-site and everybody comes
across the same problems, typically. We were
finding it very difficult to get people to
communicate to support each other. But we could
see that Sopheon's Organik could help us
there."
Organik builds
communities of interest and captures questions
presented to the community and the resulting
responses. If a question is not answered
satisfactorily it can be escalated to a human
expert. One of the other benefits of Organik is
that it provided a single point of reference for
the consultants scattered around the world. "The
consultants didn't have to think about 'who should
I ask and where should I ask'," says Sopheon's
solutions marketing director Dave Gibb. "They had
a central point to go to where they could ask
their question. If a similar problem had been
raised by a consultant somewhere else in the world
and had been answered then the knowledge base
would kick in and would give them the
answer."
In addition to
facilitating the reuse of information, Organik
also gives access to all of the other consultants
that an individual may not even know existed, what
skills they had, and so on. "You can ask
questions, find people, or you could just go in
and find someone who could help you solve your
problem and ask them directly."
Another problem
GTS-Geotech had was that, with a small staff of
highly skilled individuals, every time someone
left the company valuable knowledge left with
them. It needed some way of capturing tacit
knowledge. "We knew that there was an awful lot of
exchange of information in the background via
e-mail systems, but we weren't getting that, we
weren't seeing that," says Kaye.
"But if you deploy
something like Organik, you've captured it, it's
visible. You can see the results within the system
itself - the questions are there, the answers are
there, so your engineers feel better about
exchanging information, not having to repeat
answers."
GTS-Geotech also
detected that when Organik was deployed, its
consultants felt less insecure and isolated on the
remote site. "Organik gives them a better feeling
for the support they're getting from the whole
organisation so they know that if they run into a
problem out there, no matter what the time is,
they can get on to the web, get on to Organik, on
to our portal site, and fire off a question," says
Steve Kaye. "And nine times out of ten somebody's
awake somewhere in the world, somebody with the
knowledge and it'll come firing back."
|
Inxight also generates meta-data to define how
each piece of content has been classified. This means
that it can be automatically routed to the people who
need to access it or it can be automatically indexed in
a database to enhance future queries on the
database.
Finding Experts
Mergers not only make it difficult to find
information, they make it difficult to know where the
new experts are. "Many of the KM problems specific to
oil companies centre around being able to rapidly locate
specific engineering knowledge and employees and
subcontractors' related skills and certifications," says
Mike Tarttelin.
"If
you've bought a large company or you're merging with
another company, being able to help people make those
connections quickly can have a very important role in
the productivity of the merger," says IBM's Eric Lesser.
"You've got to help people make the introductions to one
another."
One
company that provides a solution to this problem is
Tacit Knowledge Systems. "Tacit has a special suite of
algorithms that enables you to go out and profile
everyone's e-mail and all their documents," says Peter
Auditore. "It looks at what you're working on and
automates the building of a knowledge management
system." In other words, a community of
interest.
"Texaco is automating the building of communities
of interest," explains Auditore, "because a lot of
people won't tell you what they're working on. But if
you can point the Tacit algorithms at an e-mail
repository, whether it's Notes or Exchange or whatever,
and you can look at everyone in there and get a feeling
for what people are working on, and automate the
building of a profile of what you work on on a daily
basis, then you're automating the building of a
community of interest. You don't have to ask
people."
Time
frame It's
clear that the oil industry will be facing some
difficult organisational challenges over the next few
years, from mergers, ageing experts and geographic
remoteness. The problem is not whether or not these
challenges can be met and how, but will the industry be
able to roll out the KM solutions in time.
|