Resources / Tribal knowledge
Tribal knowledge in fleet maintenance
Tribal knowledge in fleet maintenance is everything your shop knows that is not written anywhere: which unit has the sticky relay, which torque figure the manual gets wrong for your spec, which fault code is nearly always a connector. It lives in the heads of two or three people, and it leaves when they do.
What tribal knowledge means in a shop
The phrase comes from the idea of a tribe passing skills by word of mouth. In a fleet shop it means the same thing. A technician learns something the hard way, tells the person at the next bay, and from then on the shop just knows it. Nobody writes it down because everyone who needs it already has it. Until they do not.
It is different from the manual. The OEM service manual is documented knowledge. It has a page number and a revision date. Tribal knowledge is the layer on top: the local override, the pattern one person noticed across forty repairs, the reason a procedure is done in a different order here. Some of it contradicts the book, and is right to.
Two tests tell you whether a piece of knowledge is tribal. First, if the person who holds it took a month off, would the job still get done the same way? Second, could a competent technician from another yard find the answer in your files without asking anyone? If the answer to either is no, it is tribal. The longer treatment of what that layer contains is in what leaves when a veteran fleet mechanic retires.
One point of candor. Tribal knowledge is not a failure. It is how skilled trades have always worked, and a shop with none of it would be a shop with no experienced people. The problem is not that it exists. The problem is that nobody owns the copy.
Six places tribal knowledge hides in a fleet shop
Walk the shop with this list and you will find most of it inside an hour. Each entry pairs the hiding place with what it looks like when you find it.
The senior technician's memory. Fault code to root cause, across every engine family the fleet has ever run.
He hears the symptom, names the part, and is right. Nobody asked how he knew.
Handwritten notes on the manual. Margins, sticky notes, a torque value crossed out and corrected.
One annotated copy of a manual that everyone borrows. The second copy on the shelf is clean and nobody uses it.
The unit itself. A zip tie where a clip should be. A sensor relocated. A wire routed away from a heat source.
A new tech "fixes" it back to stock and the old fault returns within a week.
The parts room. Which cross reference actually fits, which supplier's part fails early, which core is worth sending back.
The parts person knows the answer. The catalog does not.
Work order free text. Years of "adjusted per usual" and "see Dave" in the notes field.
The history is there. It is not searchable, and the person it points to has left.
Group chat and phone photos. A photo of the right wiring, sent once, buried under six months of messages.
Someone says "I have a picture of that somewhere" and scrolls for ten minutes.
Notice that only the first of the six is in someone's head. The rest are physical or digital, just scattered. That matters, because it means part of the capture job is collection rather than interview. The margins, the photos, and the free text can be gathered without taking a single hour of the senior's time.
Why tribal knowledge is a risk, not just a quirk
The risk is timing. Tribal knowledge leaves on a date you do not set. The Bureau of Labor Statistics projects about 26,500 openings for diesel service technicians and mechanics each year, on average, over the decade to 2034, and most of them come from workers leaving the occupation or retiring rather than from growth (BLS Occupational Outlook Handbook, 2024). There is a fuller reading of those numbers in the diesel technician shortage, by the numbers.
Each of those departures takes a slice of the first hiding place with it. And the pattern of when capture happens makes it worse. NASA reviewed its own knowledge transfer practice in 2021 and found that processes "are retroactive after identifying the individual who is departing" (NASA, Ensuring Knowledge Continuity during Employee Transitions, 2021). One of its own respondents put it more bluntly: "there should be knowledge capture throughout the career, not after someone has given notice."
That is the same reflex in a fleet shop. Nobody records the senior's fault code memory while he is fifty. They try to in the fortnight before his farewell lunch, and a fortnight is scant time for twenty years.
There is a second, quieter risk. Some tribal knowledge is a compliance qualification. Under 49 CFR 396.25 a brake inspector has to meet a training or experience requirement, and the carrier keeps evidence of it on file. If the one person who qualifies is the one retiring, the shop has a gap on paper the day he leaves, not just a gap in skill.
A ten minute tribal knowledge check
Answer these for your own shop. Each yes is a place where knowledge is held by one person, and a place to start.
| Question | If yes |
|---|---|
| Is there a unit only one technician will touch? | That unit's quirks are tribal. Start there, it is the most tractable. |
| Does a new hire's first month include "go ask" more than "go look"? | Your documentation does not carry the shop. The people do. |
| Is there one annotated manual everyone borrows? | The margins are a capture source. Scan them this week. |
| Do work orders say "per usual" or name a person? | The free text is pointing at knowledge that was never written out. |
| Would a retirement notice from one person change your PM schedule? | You have a single point of failure and you already know the name. |
| Does your maintenance system hold the record but not the reason? | Normal. The reason is the tribal layer, and it needs its own home. |
Most shops answer yes to four or five. That is not a verdict on the shop. It is the baseline every fleet starts from, and the point of the check is to demarcate where to begin, not to score anyone.
Turning tribal knowledge into shop knowledge
The work has two halves, and they can run in parallel. The first half is collection: gather the annotated manuals, the phone photos, the group chat threads, and the work order free text into one place where they can be searched. No interviews needed. The second half is capture: short recorded sessions with the people who hold the first hiding place, one system at a time.
The pillar guide, fleet knowledge transfer, preserve it before retirement, lays out the whole approach and what to do this quarter. For the interview half specifically, how to run knowledge capture interviews with fleet veterans covers who runs them, where, and the questions that actually draw answers. If a retirement date is already set, go straight to how to capture a retiring diesel mechanic's troubleshooting knowledge, which is built around a ninety day sequence.
Where YardWise fits is the storage and the retrieval. It holds four levels of knowledge, manufacturer, fleet, engineer, and equipment, and stitches them into one cited answer. The senior's captured practice carries his name everywhere it appears. A technician asks by push to talk voice in shop talk and gets the answer back in under five seconds, with the manual page and the name of the person who set the local practice. Manuals go in as PDFs, scans, or a photo of a binder page. Captured knowledge goes in by voice, gets tagged, and gets approved by a supervisor before anyone relies on it. The fleet knowledge management page shows the whole flow.
What it does not do is schedule your PM or store your 396.3 records. Your maintenance system already does that and should keep doing it. YardWise sits next to it and holds the reason behind the record.
Tribal knowledge questions, answered short
What is tribal knowledge in fleet maintenance?
It is the shop know how that is held by people rather than written in files: unit quirks, local overrides to the manufacturer's book, diagnostic shortcuts, and the reason a procedure is done a certain way here. It sits on top of the OEM manual and sometimes corrects it.
Is tribal knowledge the same as institutional knowledge?
Close to it. Institutional knowledge is the broader term for what an organization knows through its people. Tribal knowledge is the part passed by word of mouth and never documented. In a fleet shop the two phrases usually point at the same thing.
Why is tribal knowledge a problem if the shop runs fine?
Because it leaves on a date you do not control. BLS projects about 26,500 diesel technician openings a year through 2034, most from people leaving the occupation or retiring. Every departure takes a share of the undocumented layer with it, and capture usually starts too late to get most of it.
How do I find out how much tribal knowledge my shop has?
Use the six hiding places on this page and the ten minute check. Look for units only one person touches, an annotated manual everyone borrows, work order notes that name a person, and a new hire who asks more than he looks up. Each is a marker.
What is the first step to capturing it?
Collect what already exists first: scan the annotated manuals, pull the photos out of group chats, export the work order free text. Then start short recorded sessions with the people who hold the rest, one system at a time. The pillar guide on preserving institutional fleet knowledge walks through the sequence.
See what your shop's tribal knowledge looks like once it is written down
Bring one unit and the one question your senior tech gets asked most. We will show you the captured answer with his name on it.
Book a demo