Introduction
A critical challenge of Knowledge Management (KM), from its introduction into the business lexicon, has been the dilemma of how to successfully capture, leverage and exploit tacit knowledge represented by human experience and insights. Tacit knowledge is generally defined as “know-how” — versus “know-what” (facts), “know-why” (science), or “know-who” (social).
An essential value of codifying tacit knowledge is that it can be quickly disseminated to support real-time transactions, events, and decision-making. A by-product of this dissemination is an active collaboration by stakeholders with expertise in particular domains to discuss, debate, assess, and update the new knowledge, thereby stimulating incremental knowledge and enrichment of current knowledge assets in the most effective and efficient manner.
Manage and Harness Tacit Knowledge
Over the years, KM practitioners have experienced moderate to low success in attempting to harness and manage the power of tacit knowledge with information technology. We have been witnesses to many system approaches and attempts from AI / Expert Systems to yellow pages, social software, and various specialized capture and search solutions that failed to be recognized as a sustaining tacit knowledge enabler.
The era of tacit knowledge success may be a reality with the introduction of cognitive type knowledge frameworks and systems. The cognitive-centric approach, in summary, can be systematized with the aid of digital sensors (spread across an organization or environment) that can: analyze complex situations, activities, and events; adapt quickly to evolving and differing conditions; serve as enablers to solve real business and technical problems; and generate timely decision suggestions.
A cognitive system can generally reason about causality, belief, knowledge, insights, risk, and uncertainty and are able to make decisions with incomplete data and information. By cross-referencing prior operating observations and conclusions with current activities, a cognitive system can discover and produce real-time alerts on unusual and unfathomable patterns and then quickly take action on particular conditions based on predefined rules. The rules to support the rule base are typically based on specific tacit knowledge acquired from domain subject matter experts.
Cognitive systems have the capability to generate intelligent agents and dynamically deploy them over a network. Agents fuse disparate streaming data including text, video, and graphics to provide a uniform and structured basis for problem analysis and decision-making.
Implement Approach Considerations
The cognitive approach provides for three key human touch points for infusion of tacit knowledge by domain experts:
- Populate a rules engine with predefined rules applied to incoming events that trigger the detection of patterns or trends of significance in real time.
- Update rules engine with new knowledge as it becomes available so as to tweak the system, augment the rules and increase the likelihood of achieving predefined goals.
- Interact with the system on an ongoing basis and test system-generated alerts and suggestions with human experience and insights.
Real Life Tacit Knowledge Experience
The Deepwater Horizon oil rig explosion, some years ago, that killed workers and poured millions of gallons of oil into the Gulf of Mexico is a good example of how tacit knowledge can be employed in real-life situations. Considering the fact that offshore oil rigs are comprised of an array of complex integrated subsystems and that their use is controlled by many software applications and technical environments, is it possible a software failure could have contributed to this tragedy? Is it also possible that the tacit knowledge touch points (as outlined above) were either not a consideration, not implemented, or not observed as a monitoring tool as required by current standards to ensure correct and timely alerts and decision-making for system failures?
The Gulf disaster underscores the fact that sometimes complex environmental events happen too fast for appropriate system discovery, analysis, and reporting. It also shows that human decision-making based on sensed data often isn’t rational and may underestimate associated risks. Is the inclusion of a final fail-safe review activity by a domain expert required when a complex event factor enters a predefined danger red zone?
In conclusion, I believe an appropriate quote to frame this article is one of my favorite Deming quotes as shown below.
“Best efforts will not substitute for knowledge.”
Dr. W. Edward Deming
