Network Operations · OT/SCADA · Applied ML

Engineering at the edge of
the power grid and machine learning.

I'm Josué Aldana — a network operations analyst building data pipelines, protection-coordination tooling, and ML systems for electrical utilities. From a fleet of 600+ field devices to high-altitude balloons, I turn raw telemetry into decisions.

Field devices managed
0+
Network nodes modeled
0
IEEE publications
0
Feeders & substations
0 / 31

Selected projects

Production tooling and research builds across grid operations, data engineering, and embedded systems.

RAG · NLP In production

SOP Knowledge Retrieval

Retrieval-augmented pipeline over ~3,150 historical switching-order documents, with multilingual embeddings, vector search, and phase-by-phase validation for grid operators.

  • ChromaDB
  • e5-large
  • Claude
  • Python
RF · GA Published

rf-evolve

Genetic-algorithm optimizer for SCADA concentrator placement using Longley-Rice/ITM propagation and Fresnel modeling over national SRTM terrain. Paper submitted to CONCAPAN 2026.

  • Longley-Rice
  • SRTM
  • GA
  • KML
DMS · Graph Internal tool

DMS Topology Engine

Pipeline processing 14 ADMS tables into per-feeder network models — ~190k nodes, 228 feeders, 31 substations — with a Streamlit dashboard and HMI-style single-line diagrams.

  • NetworkX
  • Parquet
  • Streamlit
  • SVG
Protection Validated

Protection Settings Consolidator

Multi-source consolidation of relay settings across SCADA RTDB, engineering forms, and ASPEN — a 432-equipment catalog with HV/LV normalization and 11 automated coherence checks.

  • ASPEN
  • DNP3
  • pandas
  • IEC 61850
HAB · Embedded Active

eohab2026

High-altitude balloon research integrating the CUSF burst calculator with SondeHub Tawhiri trajectory prediction, plus ESP32-S3 edge payloads under the scitechlab-pi label.

  • ESP32-S3
  • Tawhiri API
  • Python
  • TinyML
BI · Fabric In progress

IOP Data Architecture

End-to-end BI architecture for an operations unit: SharePoint → Microsoft Fabric notebooks → Power BI semantic model, with ETL pipelines documented as living diagrams.

  • MS Fabric
  • Power BI
  • SQL
  • ETL
Josué Aldana
Josué Aldana Network & ML Systems Engineer · San Salvador

Where grid operations meet data science

I'm a Network Operations Analyst within the Operations Engineering unit at an electrical distribution utility, working across protection coordination, SCADA/OT systems (DNP3, ICCP, Modbus), DMS topology modeling, and data engineering pipelines for a fleet of 600+ intelligent electronic devices.

I also teach electrical engineering as a part-time instructor, and I'm currently pursuing a master's in data science. My foundation is OT/SCADA and protection systems; my trajectory is toward solutions architecture and ML systems for the energy sector.

Focus OT/SCADA · Protection · Data Engineering · Applied ML
Toolbox Python · OpenDSS · ASPEN · ChromaDB · Power BI · ESP32
Standards IEC 61850 · IEC 62443 · NERC-CIP
Recognition 14 IEEE publications · best graduate · ICTP & MIT-adjacent training

Publications & research interests

14 peer-reviewed IEEE publications. Ongoing work in adaptive protection, fault location, and physics-informed learning for power systems.

Current research lines

  • Adaptive seasonal protection Unsupervised clustering + probabilistic deep learning + supervised regime classification, validated in OpenDSS.
  • Fault location via multimodal fusion Physics-informed learning combining oscillography (COMTRADE) with topology and telemetry signals.
  • Network optimization with GAs Genetic algorithms applied to RF concentrator placement and grid communications planning.

Profiles

Topics span COMTRADE oscillography analysis, power-systems simulation (OpenDSS, ASPEN), IEC 61850/62443, and embedded systems / IoT.

Let's build something

Open to collaboration on grid data systems, applied ML for energy, and research. The fastest way to reach me is below.