Chasing the vapour trail

There are times when tacit knowledge isn’t good enough. What if your product life cycle is at least 40 years and in some cases it’s still in use after half a century? You stopped production 30 years ago and all of the original designers and service technicians have long since retired or passed. There is still a strong demand for parts and for knowledge long after production ceases. Customers need knowledge to assist in problem-solving. Most importantly, human life depends on it.
William Boeing began building airplanes in 1916 with Bluebill, a utility seaplane which was sold to the New Zealand Flying School and later used for New Zealand express and airmail delivery.
Boeing’s planes were at the vanguard of the development of commercial air travel and the aircraft were pressed into service in both World Wars I and II. Just recently Boeing received a technical service query for an A-20, an airplane that served in World War II. Amazingly, the knowledge requested was available and provided to the customer.
It might have been otherwise.
With the advent of technology, the speed of business and changing demographics of a technical workforce, Boeing has had to re-address how it manages knowledge: more importantly, how to capture it and make it explicit so that its life cycle transcends that of the designers and technicians.

From ‘Masters’ to ‘Padawans’
Drawing on the language of Star Wars, Paulette DeGard, knowledge strategist for Flight Deck Engineering at Boeing, says its knowledge workers have gradually gone from long-term “Masters” to nearer-term “Padawans” (apprentices or learners in the language of the Jedi).
“Suddenly the issue of making tribal knowledge explicit becomes grossly obvious,” DeGard says, “and quite difficult due to the long time that has elapsed since much of the knowledge was generated.”
Hunting for elusive knowledge has become a time-consuming quest at Boeing. Long ago (at the end of the last century), a Gartner Group study on the use of knowledge workers’ time concluded that 60 percent of their time was spent looking for information to improve their knowledge and work. At Boeing, that would not have related to the new Padawans, but to the old Masters.
More recently, a report by Douglas Holtz-Eakin, director of the Centre for Geoeconomic Studies, Washington, DC, US, on the prospects of Baby Boomer retirement showed that 70 percent of the working population would be eligible for retirement by the year just ending – 2007.These prospects portended dire days ahead for Boeing.
“So how does a company like Boeing address a situation of this magnitude?” DeGard asks. “We eat the elephant one bite at a time.”
More than two years ago the Flight Deck Engineering organisation determined it needed to make a concentrated effort to address the changing demographic of the group. ‘Greybeards’ were retiring and young technical workers were replacing them.
They approached the problem like engineers but they were not suited to creating a knowledge management system. And so they hired Paulette DeGard, previously an external business analyst for the company with considerable KM experience.
“The first mandate I was given was to design a system that a new Boeing engineer 10 years from now could use to find the information they needed and know that it was the right information,” DeGard recalls. “The second mandate was that this new employee could bet their pay cheque on it.”

Capturing diffused knowledge
Quickly she discovered the crux of the problem: what she calls “individualism diffusion.” The advent of personal computers had diffused knowledge and information into individual databases, each structured differently and without agreed upon hierarchy or file names, formats or forms. In this condition it was utterly impossible to share the existing explicit knowledge.
The solution was for the Flight Deck group to identify an agreed upon taxonomy. “Otherwise,” DeGard warned, “we will continue to dumpster dive servers just like today.”
Several repositories (stand alone servers) were set up to capture previous design decisions, graphics, working knowledge and presentations. These repositories would have varying levels of confidence.
For the previous design decisions, an approval board made up of senior engineers and managers would be used to validate the correctness of the information and would then have “read only” access. The graphics and presentations repositories would be medium confidence with “read and write” access allowing people to place material into the folder at will. The working knowledge repository would also be medium confidence and dependent on individual input.
With the taxonomy in hand, Flight Crew Operations Integration (FCOI), a sub-group of Flight Deck Engineering, was identified as a pilot project group. From the outset of the test, it was determined FCOI would become a “learning organisation” by shifting from individual-based learning/organising to organisational-based learning/organising.
“The group agreed that reliable knowledge management could reduce mistakes related to inconsistent design decisions,” DeGard recalls. And, the group agreed that would require reliable and repeatable processes to capture, deploy and ensure accuracy.

Building on commonalities
Then the hard work began. DeGard interviewed some of the engineers and learnt they were all organised differently, even though many used Microsoft Outlook. Then she asked every engineer in the Flight Crew unit to send her a screen capture of their folders.
From the folders she tried to identify commonalities and discovered they actually did organise similarly at an aggregate level.
In the aviation industry there is a standard for categorising parts – the Airline Transport Association (ATA) chapters.
DeGard discovered the majority of the engineers (unaware of how the others were doing it) were organising their knowledge and information either by ATA or Boeing models – most by model first, then ATA chapters. She developed her first taxonomy model around the majority practice.
The next issue was how to assure accuracy. “At Boeing, like many corporations,” DeGard says, “people believe what they read because they know the author and know he or she is qualified. This personalisation strategy works as long as you know the person. But to bet your pay cheque on the validity meant we had to move from a personalisation strategy to a process that validated the accuracy of the documents.”
To validate every single document was daunting. And so, the tiered system of repositories, representing varying levels of confidence, was launched. Documents that contained design decisions with impact on the future were reviewed and placed in the folder at the highest confidence level.
The final factor was assuring these repositories would be able to stand the test of time, especially business cycles in production and workforce. How would they survive downturns that resulted in budget cuts and force reduction?
“The solution was to make certain the repositories had an economic value that would validate the need for maintenance,” DeGard says. Now, when documents are identified for placement in a repository, a monetary value is associated with them. We use a formula of eight hours per page times the rate scale for the engineering involved.”
More issues developed as the project moved forward.
Chief among them was finding documents out in the larger world of servers, emails and filing cabinets. “It became quite evident that a lot of the information from the past 10 years was resident in emails – conversations, threads and attached documents,” DeGard recalls.

Capturing floating knowledge
She searched the Internet for tools that search and organise emails and found the Email Intelligence Platform™ by Clearwell Systems.
“This system allows you to feed Outlook files into the system, which then crawls and indexes the emails,” DeGard says.
“The benefits of the system are to be able to rapidly capture, and accurately categorise Flight Deck knowledge found in emails and their attachments, provide simple and intuitive access to Flight Deck knowledge for existing and new employees, accomplish KM capture in a cost effective and efficient manner, and establish repeatable process for future KM capture needs.”
Bottom line: Clearwell discovers, organises, and analyses the information captured in emails.
To test the email system, a new employee, who had little technical knowledge, was asked to find the answer to a highly technical question using captured email. In less than 45 minutes he was able to identify the information and answer the question. Additionally, he located documents valued at over $500,000 USD which came from actual test flight documents validating the answer.
“What we have discovered,” DeGard reports, “is that the use of formal decision making documents and processes has become more informal through the use of email. In the past there were meetings and formal events that constituted a design decision or some other important milestone. Today, many of these decisions are agreed upon in emails and not formally documented in the same way.”
Like other companies, Boeing has moved from formal structured to informal unstructured decision processes over time. DeGard believes this shift has been good because it speeds up the process. However, it is more difficult to locate specific decisions and ensure appropriate stakeholders are included unless they are actually involved in the email chain.
“The Clearwell System is allowing us to capture and store these informal unstructured decisions, which will allow employees today and in the future to have quick, accurate access to vital information for accomplishing their jobs,” DeGard says.
Other benefits of the system include the ability to use Boeing taxonomy in organising email and therefore using a consistent classification technique. “Probably the best feature we have found is that when the system indexes the emails, it removes all duplication of content. This is extremely valuable for being able to read through threaded conversations and reduce the amount of ‘noise’ you typically have to sort through.”

Assuring privacy, security
An equally important factor is that Boeing has many global partners, customers and employees and it has major concerns about privacy, intellectual property retention and security of information. During the pilot project, architects have had to look closely at factors like International Trade and Arms Regulations (ITRs) as well as issues of proprietary information.
DeGard recognises, “These issues have to be resolved as we move from an offline pilot to a more accessible system.”
To understand these non-technical problems, she has identified concerns from various groups: Information Protection, IT, Legal, Procurement, Records and Information Management, and Email IT architects.
But on the positive front, there are several major company-wide initiatives with which the KM project fits.
“We are being asked to adopt Lean Engineering as a corporation,” DeGard says. “This is a huge effort and one that impacts everyone in the company. For us in Flight Deck this means that we need to look closely at our processes and deliverables and identify ways of eliminating waste. In the engineering world, waste is defined as re-work or queue time waiting for someone to make a decision or input.”
The Flight Crew pilot project is addressing that very task by re-working processes, capturing knowledge and information and reducing costly search time and waste.

From takeoff to flight
The pilot project was completed in December 2006 having made good progress in refining the methodology and toolkit.
“We continued to work on the approval process, which now works well with a series of questions that leads to the level of confidence we can have in each document,” DeGard says.
In May the team began the next phase, expanding out to all of the Flight Deck, Environmental Controls and Flight Operations Test Pilots – from 25 people to over 600 in the course of a few months.
“To this point,” DeGard notes, “everything is working as expected. The new groups have helped us clarify a few things, making the toolkit and methodology that much stronger.”

Conclusion
Managing knowledge is a critical and daunting task regardless of the variety, but locating and capturing missing knowledge and making it available for reuse is perhaps the most challenging of all, particularly where the shelf life of knowledge is long and the turnover of ‘knowers’ is short. The young Boeing system shows great promise and – with continued leadership and financial support – could stand the test of time.

Jerry Ash is managing editor, Inside Knowledge, KM coach, and founder of the Association of Knowledgework, http://www.kwork.org. He is author of the Ark Group’s Next Generation Knowledge Management series. To order any of the three volumes, contact Adam Scrimshire at publishing@marketing.ark-group.com. Jerry Ash can be reached at jash@kwork.org.

For further information on the Boeing programme, contact: paulette.h.degard@boeing.com.