Case study · Utilities
A week of diagnosis,done in minutes.
One of the world’s largest electric utilities and a leader in ultra-high-voltage direct current transmission, operating a transmission network that spans millions of kilometers. It manages critical energy infrastructure including converter transformers, the core UHVDC equipment, each built from more than 1,000 components. The sheer volume and criticality of that equipment, and the risk of catastrophic chain failures from an undetected fault, is what required an AI assessment system. Manual diagnosis previously took more than a week per fault.
Constraints and answers
Diagnosis by experience does not scale, and does not repeat.
The system that made diagnosis repeatable
Fault positioning followed experience, not a standard
Experience-based fault positioning with inconsistent standards led to deviations in diagnosis.
Answer
Intelligent assessment agent
Trained on a large corpus of historical fault cases, applying one standard diagnostic chain of monitoring, diagnosis and reasoning across all common fault types.
More than a week of manual sorting
Manually sorting 8 data categories and 154 data points took more than a week per fault.
Answer
Automated data ingestion and analysis
Reads all 8 equipment data categories and 154 monitoring points and produces structured assessment reports without manual sorting.
No decision record to reuse later
Without systematic decision records, experience could not be reused on similar faults.
Answer
Knowledge management system
Traceable decision logic capturing the full diagnostic pathway, from description through cases to guidance, so knowledge is retained and reused.
No automated trace from fault to decision
With no automated analysis tools, faults could not be traced quickly or decisions issued.
Answer
Large-model reasoning pipeline
Sensory detection inputs combined with AI reasoning, replacing the manual analysis workflow and speeding up fault tracing.
Every fault runs the same chain, and every step of it can be read back.
The results
From a week to minutes, on 2,000 learned cases.
- 2,000
- HISTORICAL CASES LEARNED
- 13 common fault types diagnosed on one standard chain, with experience reuse 300% faster.
- 95%
- BETTER ANALYSIS EFFICIENCY
- Higher diagnostic accuracy and lower maintenance cost, with steadier equipment operation.
- FAULT ASSESSMENT TIME
- Down from over one week to minutes per fault assessment.
- POWER SUPPLY RELIABILITY
- Stable equipment operation means a safer supply across the network.
An assessment an engineer cannot audit is an assessment an engineer will not act on.
The stack
What was deployed
Automated ingestion and analysis of 8 equipment data categories and 154 monitoring points for a complete fault assessment.
Systematic capture and reuse of diagnostic decision logic and historical fault cases, so the institution keeps what it learns.
Enterprise-grade controls protecting critical infrastructure data and holding regulatory compliance.
Traceable diagnostic pathways, from description through cases to guidance, giving technical teams clear reasoning to check.
Note on Capabilities
The capabilities, credentials and deployment references shown across this site reflect the collective work of Amanah's core engineering team. Certain engagements were delivered under prior or affiliated entities. Sovereign deployments are confidential by mandate, so individual customers are not named. The underlying references are verifiable under NDA.
Request a Strategy Session
The same platform, inside your borders.
Book a working session with our sovereign-architecture team. We map your mandate, residency and audit requirements to a deployment plan your own people own and operate.
We reply within one business day. Your inquiry stays with our team, never a third party.