What if the fault tree could talk to the service technician?

Expert knowledge no longer locked in a few heads

At many manufacturers, technical service depends on a handful of experts. They know the fault patterns, the readings and the order in which to check things. Service technicians and dealers call, wait for a call back, and the experts rarely get to their actual work.

The knowledge is often already documented in fault trees, repair guides and product data sheets. It’s just hard to get to when you’re standing in front of a machine with a multimeter.

From symptom to solution

• Channel
• Project tool
• Skill
• LIVOI tool

1. 1

   Channel Describe the symptom

   The technician describes the problem, even on the road via WhatsApp with a photo.

2. 2

   Project tool Find the fault tree

   Visio fault trees were converted into searchable text, including yes/no branches.

   search_fault_trees
3. 3

   Skill Step by step

   The agent asks the next check question and follows the matching branch.

4. 4

   LIVOI tool Guide & parts

   Repair guide and product info from knowledge search, with source.

   rag_search_tool

The fault trees are still maintained in Visio. A tool built in the project converts them again on every upload.

In practice: a generator manufacturer

A manufacturer of generators and mobile power and propulsion systems has used LIVOI to make its technical knowledge centrally available, from fault trees and repair guides to detailed product information. Today, the agent answers most technical questions on its own or guides technicians step by step through the repair. Service handling times have dropped by 30-60% as a result.

Fault trees as a tool

A project tool reads Visio diagrams, converts nodes and connections, including labels such as “yes” and “no”, into text and makes them searchable.

Access for authorized users only

A whitelist gives service technicians and authorized dealers access, for example via WhatsApp.

Experts for the hard cases

The agent handles standard questions itself. If it gets stuck, the expert team takes over the chat with the full history.

Knowledge stays current

New guides and updated fault trees are uploaded and available immediately.

15-second demo: diagnosis in conversation

Note

The following chat messages are a simplified example. The fault, check steps and values are made up.

Service technician

Generator starts but shuts down after about 10 seconds. No error on the display.

Project toolsearch_fault_trees

Fault tree “Engine stops after start”

Service agent

There’s a fault tree for this. First check: Is there sufficient fuel at the fuel filter? Please answer yes or no.

Service technician

Yes.

Service agent

Continuing on the “fuel present” branch: Is there continuity at the oil pressure switch while the engine is running?

Service technician

No.

LIVOI toolrag_search_tool

Repair guide oil pressure switch, section 4

Service agent

According to the fault tree, the oil pressure switch is probably faulty and shutting the engine down. Replacement is described in the repair guide, section 4. Please check the oil level first.

Making knowledge usable, not just uploading it

Fault trees and guides alone don’t make a good service agent. Together with the service team, we clarify how questions are handled today, when an expert should take over and how new knowledge will be maintained. More on our approach

What changes day to day

Before

• Technical hotline permanently overloaded
• Knowledge in a few heads and scattered diagrams
• Technicians wait on site for a call back

After with LIVOI

• Guided diagnosis around the clock
• Fault trees and guides in one place, with sources
• Experts focus on the complex cases