Carlos Leyva

Silicon Stories

Chapter 8: Knowledge Management & Infotainment

Big Automation Company

« PreviousContentsNext »

After the dot.com implosion I have spent a significant amount of my time working with Big Automation Company on their next generation simulation product. They have a 15-year track record of providing high fidelity training simulators to some of the largest oil refineries on the planet. The Distributed Control Systems (DCS’s) that run these refineries are 24/7 real-time systems with NASA like high availability requirements (i.e. they almost never go down). Anytime these systems are down for even short periods of time (i.e. minutes) the refineries lose millions.

The fact that these systems run 24/7 is great for refinery productivity but makes the training of refinery operators a difficult task. To address this problem the refineries have been using high fidelity simulation based training software since the advent of the DCS’s. The simulation software contains graphical representations (i.e. models) of sections of the refinery that instructors use to develop training scenarios. These scenarios include: plant startup and shutdown, equipment malfunctions, DCS alarms, etc. Since the simulation models contain accurate virtual representations of plant processes and equipment they provide an excellent platform for the development of context sensitive rich content.

To my knowledge none of the simulator-based training systems, including those from Big Automation Company, have the functionality I describe in the following text, but they could. The technologies required to implement these ideas are all commercially available today. As mentioned, the plant models used in simulator training systems for refineries contain graphical representations of real world plant objects such as pumps, valves, distillation towers, etc. In addition, the models contain a representation of the plant topology. That is, how all the pieces of equipment are “laid out” and connected to each other.

Imagine clicking on a pump icon and selecting a menu option that describes the maintenance history of that specific pump. In addition, you can select another menu option that plays an audio/video clip describing the procedure required to maintain this pump to the specifications proposed by the manufacturer. The man in the video is one of the senior operators from this section of the plant.

Need additional information regarding this pump? Select another menu option that takes you to the manufacturer’s web site where you can talk via real time chat with a manufacturer’s representative about recent problems that you experienced with this piece of equipment. Sounds far fetched? Perhaps it is, but only because we are unaccustomed to thinking of computing systems along these lines. This idea is much too obvious not to find its way into leading edge products within the next few years. It is a killer app in the making!

I can hear the geek wheels turning. Leyva, you dumb shit, who could build such a system? The automation vendors that supply the simulation software would need to collaborate with perhaps hundreds, if not thousands of manufactures in order to pull this off. And what will it cost to produce maintenance videos for every piece of plant equipment? Well I never said that it would be trivial or inexpensive (big assed grin)!

Bare with me a moment and lets explore the possibilities. Web services are, as we speak, standardizing app-to-app communication over the Net; so communicating with a large number of manufacturers will soon be plausible (notice I didn’t say easy). Producing audio and video clips for every piece of maintenance equipment would be time consuming, but the technology to produce it and persist it is available today, and getting cheaper all the time. Besides, how expensive will it be when blue-collar boomers retire and take all their tacit knowledge out of the plant with them?

Lets assume, for the sake of argument, that all this is not only doable, but cost effective as well. Why would it be important vis-à-vis KM? Because if context sensitive rich content can lead to real time conversations then one of the missing pieces has just been added to the KM puzzle, people. Human beings conversing with other human beings lead to relationships and the building of trust. Once trust is established human beings begin to engage in one of the things they do best, telling stories. Stories lead to the effective transfer of information, which leads to experimentation, which leads (possibly) to the accumulation of new knowledge.

« PreviousContentsNext »