Multimedia formats facilitate actionable information
As seen in these screens, Pensieve employs multimedia file formats to support the best medium for conveying information
Unlike current solutions, Pensive is easily editable, which allows current annotations to stay up-to-date
Our research shows that technicians utilize multiple factors to determine the reliability of the annotations. We discovered that the trustworthiness of an annotation depends on when it was last updated, who wrote it, how many people validated it, and the annotation's relationship to other sources of information.
One of our key research findings was that veteran technicians possess, and active technicians collect, crucial information and knowledge on the efficacy of work execution procedures. However, consulting a technician in the authoring of a WAD document is neither mandatory nor institutionally reinforced. “Good” WAD authors consult technicians on the clarity and efficacy of a WAD document, while others forge ahead without doing so.
A technician is not expected to know everything which an engineer is responsible for, such as in understanding the needs or reasons for why a work execution procedure might be conducted the way that it is. But what we do know is that stronger, more effectively written WADs consult technicians and incorporate feedback from the people actually performing the work. Technicians are “front-line data collectors” for what works best, least, or not at all in the execution of a given WAD. Therefore, technicians maintain highly valuable, though under-utilized knowledge about work execution.
Therefore, the consequences of incorporating and embedding technician knowledge within a WAD execution interface include work instructions that are more clear and therefore more efficiently or rapidly executed and a reduction in errors and mistakes due to a lack of comprehension or understanding.
The information needs for technicians executing WAD documents are therefore twofold.
Technicians need: a) existing reference information and documentation (hardware manuals, design details from the OEM, or training resources for technicians and procedures) which, to date, remain unintegrated or inaccessible from the WAD itself, and b) access to “tribal” or common knowledge which is either implicitly or tacitly known by experienced technicians with a prior understanding of the work procedure or system.
Our product leverages these two information types within a web-based interface to surface relevant information to technicians executing WADs. The solution we propose integrates existing information resources to improve technician work execution, in addition to surfacing “tribal” or common knowledge during work execution.
Knowledge management (KM) is a huge undertaking at NASA, with programs and initiatives aiming to educate, distribute, and disseminate knowledge in matters of high-level systems-based integration, but also design engineering processes, as well as decision-making from a management perspective.
Knowledge management encompasses not only the dissemination, sharing, and networking of knowledge resources across NASA agencies, but also the development and expansion of knowledge resources through active documentation and/or codification.
KM systems not only capture, store and develop engineering knowledge and information within databases and repositories, but also retrieve, surface, and “mine” that knowledge for users.
At NASA, some of this knowledge includes:
In an ideal scenario, when a WAD procedure is executed, the information gleaned from performing the work ought to continually improve the execution of the procedure in the future—to either prevent problems and alterations that were encountered or to improve the speed and efficiency of the work that was executed. Therefore, acquiring that knowledge is just as important and critical as its communication.
Other Knowledge Management initiatives across NASA and the aerospace industry include:
Our research has demonstrated that Solumina and Maximo do not yet fully integrate with design engineering drawings or materials. Very often TOSC workflows are barred from accessing third-party OEM engineering documents available to NASA engineers, as an example. In an ideal scenario, design information for the hardware assembly (drawings and specifications), training materials for procedures, or manuals relevant to the operation of a hardware element would be available for a technician to consult on-the-fly.
As knowledge management resources are further developed to document and capture technician experience during work execution, knowledge resources must more effectively integrate with work execution workflows to empower technicians to solve problems just-in-time, anticipate potential issues, and mitigate risks. Information that would be found within existing Lessons Learned repositories, Best Practices repositories, as well as video lessons or tutorials created through programs within NASA’s Critical Knowledge Gateway would integrate with the content of the work authorization document itself.
Engineering, manufacturing, or fabrication knowledge is only as good or useful as it’s capacity to be surfaced at the right or relevant time. The problem with specialist knowledge (such as that possessed by veteran technicians) is that it remains undocumented. Institutionally, this is a major problem and barrier to long-term growth.
With an aging aerospace workforce, retention of that knowledge is of utmost importance. When older generations retire they take their knowledge (lessons learned, time-saving tricks and optimizations included) with them—so a retiring workforce also means retiring knowledge. If you have one long-term employee who is the only person who knows the specifics of how to a number of procedures, what happens when that person leaves? If the information isn’t documented then no one knows about it.
A major finding of our research is that even when ArcJet technicians and engineers encounter a crucial lesson learned, there is little incentive to document that finding within their existing log or database. To make matters worse, there is even less incentive to consult or refer to that log or database when testing or work procedures are conducted and performed.
Knowledge management initiatives strive to document veteran technician knowledge through procedure documentation and training materials. We see our product leveraging these resources to provide both new and experienced technicians the knowledge base of an entire user base.
The knowledge base relevant for work execution and other technician responsibilities is inevitably built through experience.
“Tribal knowledge is any trick, inside tip, shop aid, or bit of experience that helps someone produce a part,” Agne explains. “When you outsource, the supplier gets the drawings, the planning papers, and the tooling to make the part. But unfortunately, he doesn’t know about, for example, a shop aid that is critical to making the part. The only person who knows that is the guy that has been making it for 17 years. So there’s a learning curve involved. There’s a certain amount of knowledge that’s not reflected on the planning documents. It’s something that is lost and has to be regained, and we try to make suppliers aware of that knowledge before they start making parts.” (Chuck Agne, Director of Supplier Management for Boeing’s Integrated Defense Systems (IDS)
"I've learned more from people with dirty fingers than I've ever learned from those in the classroom."
What types of knowledge do experienced technicians possess about work execution procedures?
“One enhancement was simple but effective: Place dividers between individual parts in totes, and add an inventory sheet to the front, rather than jumble them together. This segregation made it easier to complete a parts accounting before the totes were delivered to the factory floor.”
While our prototype for Pensieve focuses on surfacing relevant information within work execution workflows, the documentation process is just as important within the product’s high level development.
The first phase of the product’s adoption will focus on documentation and capturing of relevant work execution knowledge. As more users adopt the crowd-sourced platform, more data will be created on work execution procedures, strengthening the quality of the product. Likewise, as the product is iterated and develops in the future, the integration and cross-referencing of data with other knowledge management repositories will strengthen knowledge management across NASA facilities.
Once sufficient data has been captured, Pensieve will mutually strengthen other knowledge management platforms through cross referencing of information. Some of these initiatives or complementary products have been previously implemented as long-term goals at Kennedy and other NASA centers alike.
By focusing on the process of knowledge capturing first, and the transmission and distribution of that knowledge to our users, we will provide a product of great value to benefit the efficiency, reliability, and safety of technician work execution processes. The result will be a companion technology leveraging the wisdom of an entire community’s knowledge base, portable and easily referenced while on the job.