Maintenance Knowledge Assistant

Cited answers from manuals and service bulletins for technicians.

Industry	Level	Difficulty	Recommended mode
Aerospace	1 of 5	Beginner	flat

Coming soon

A complete, runnable Maintenance Knowledge Assistant reference implementation is in progress. The architecture and core code below already work with SrvDB 0.3.

The problem

• Manual sets are huge.
• Bulletins update procedures.
• Hangars may be offline.

How SrvDB powers Maintenance Knowledge Assistant

1. Manuals and bulletins are embedded by task.
2. Questions return cited tasks.
3. Runs offline on tablets.
4. Revisions replace old content.

What changes

Today	With Maintenance Knowledge Assistant on SrvDB
Searching manuals	Cited tasks
Missed bulletins	Retrieved together
Connectivity needed	Offline

How it works

1. Ingest current revisions.
2. Embed.
3. Answer with citations.
4. Log usage.
5. Update per revision.

SrvDB features it uses

• Exact search (flat): every query compares against every stored vector, so results are exact and explainable.
• upsert: inserts new items and replaces changed ones, so the knowledge base stays current.
• Fully offline: SrvDB runs in-process and never opens a network connection.
• Crash-safe persist(): after it returns, data survives a crash or power loss and reopens exactly as committed.

Use cases

• Task lookup
• Torque values
• Bulletin applicability
• Tooling
• Inspection intervals
• Training

Core code

embed() is your embedding model (for example sentence-transformers, or a function that turns your records into feature vectors).

import json
import srvdb

db = srvdb.Database("./maintenance-knowledge-assistant-db", dimension=768)

def ingest(items):
    """items: list of (id, text, metadata dict)"""
    db.upsert(
        ids=[i for i, _, _ in items],
        embeddings=[embed(t) for _, t, _ in items],
        metadatas=[json.dumps(m) for _, _, m in items],
    )
    db.persist()                      # crash-safe checkpoint

def find_tasks(question, k=5):
    hits = db.search(embed(question), k)
    return [(json.loads(db.get(i)), round(score, 3)) for i, score in hits]

Limits and honest notes

• Official manuals remain authoritative.

Caution

SrvDB is not certified software. Use it for decision support and knowledge retrieval, never in flight-critical or certified systems.

Build roadmap

1. Prototype: load a sample of real data with flat mode and check that searches return what an expert expects.
2. Measure: if you move to an approximate mode, compare its results with flat on the same queries (recall).
3. Pipeline: add continuous upsert and decide when to call persist().
4. Harden: test restarts and crashes, add logging, review every output path with domain experts.
5. Deploy: ship the database directory with the application or service.

All Aerospace projects

Level	Project	What it does
1 · Beginner	Maintenance Knowledge Assistant (this page)	Cited answers from manuals and service bulletins for technicians.
2 · Intermediate	Safety Report Analyzer	Clusters safety reports and links similar events.
3 · Advanced	Fleet Finding Matcher	Matches new maintenance findings to similar findings across the fleet.
4 · Expert	Borescope Image Search	Find similar borescope images and the findings they led to.
5 · Frontier	Mission Knowledge System	Offline mission and ground-segment knowledge for space operations.

See also: Use cases · Live examples (coming soon)