Fault Identification
Parse and classify the incoming fault code — SPN/FMI, OBD-II, or natural language description.
AI Diesel Tech is not a chatbot. It's a deterministic finite state machine diagnostic engine — built by ASE Certified Diesel Technicians to think the way a master technician thinks.
A Finite State Machine (FSM) is a computational model that exists in exactly one of a finite number of states at any given time. Transitions between states are triggered by specific inputs — in this case, fault codes, test results, and technician observations.
The word deterministic means that given the same inputs, the engine always produces the same output. There is no randomness, no hallucination, no creative interpretation. This is fundamentally different from large language models that generate probabilistic responses.
For diesel diagnostics — where a wrong answer costs real money and can cause safety failures — determinism is not optional. It's the foundation.
Every diagnosis progresses through these states in sequence.
Parse and classify the incoming fault code — SPN/FMI, OBD-II, or natural language description.
Cross-reference reported symptoms against the knowledge base to build a probability-weighted fault hypothesis.
Generate an ordered sequence of physical tests — cheapest and most likely first — to confirm or eliminate each hypothesis.
Evaluate each test result against expected values. Narrow the fault tree based on pass/fail outcomes.
When sufficient evidence is gathered, lock in the root cause with a confidence score and supporting evidence chain.
Output a prioritized repair plan with part numbers, torque specs, and verification steps.
Special-case handlers that activate when standard state progression is insufficient.
Activates when a fault is present in history but not currently active. Adjusts test sequencing for intermittent conditions.
Detects when multiple related fault codes are present and resolves the root cause before addressing downstream codes.
When test results are ambiguous, the engine requests additional data or escalates to the Technician Chat™ module.
For safety-critical faults (brake, steering, fire risk), the engine immediately surfaces the highest-priority action.
Six technical differentiators that separate the FSM engine from generic AI tools.
Generic AI models generate probabilistic responses — they guess based on patterns. The FSM engine follows a deterministic logic path. Given the same inputs, it always produces the same output. This is critical for safety-critical diagnostic work.
Every diagnostic session navigates a structured fault tree built from real OEM diagnostic logic. The engine never skips steps, never hallucinates part numbers, and never recommends a repair it cannot justify with test evidence.
The Technician Thought Module (TTM) is an internal reasoning layer that evaluates test results in context — considering vehicle history, mileage, environmental conditions, and prior repairs before advancing to the next state.
At each diagnostic step, the engine injects relevant reference data — wiring diagrams, connector pinouts, component locations, TSB matches — directly into the technician's view. No searching. No switching tabs.
Technical Service Bulletins are scored against the active fault using a weighted algorithm: SPN/FMI match (40 pts), VIN range (30 pts), year/make/model (20 pts), engine platform (15 pts), keywords (5 pts). Top matches surface automatically.
Diagnostic facts are stored in a normalized 16-table relational schema — not a flat knowledge base. This enables precise retrieval, cross-referencing, and continuous enrichment without data degradation.
The best way to understand the FSM architecture is to run a diagnosis. Pick a plan and start diagnosing today.