Hi, I'm Anton Andreenko — a an Engineer with over 17 years of experience.
I hold a Master’s degree in Computer Science, where I was first introduced to artificial intelligence and neural networks.
I see plenty of opportunities for AI implementation in controls automation:
-
Quality control
-
Anomaly Detection
-
Network security
-
Alarm/Log management
-
Troubleshooting
...and many others.
The rise of foundation AI models has given a huge boost to this direction, making experimentation easier and more accessible than ever before.
My current work explores:
AI solutions for predictive maintenance, anomaly detection, and process optimization
Integrating AI into enterprise environments
Developing intelligent solutions to enhance safety, efficiency, and decision-making
A sample script that shows how Python can be used in Alarms Management. It loads alarms from csv file and calculates varisou metrics (per ISA-18.2 and EEMUA 191 guidelines) and also generates recommendations. All calculations are extrame inexpensive and can be done on any computer (and even mobile phone).
Github: https://github.com/andreenko-repo/alarms_analyzer
A small tool that extracts logic and variables from Studio 5000 L5X exports. Extracted informtion is much mor ecompact than original L5X file, so it can be more effciantly used for analysis using AI NLP tools.
Github: https://github.com/andreenko-repo/rockwell_logic_l5x_extractor
An experiment (inspired by Google and DeepMind SIMA-2) using a Gemini-based AI Agent to control the Tennessee Eastman Process through HMI screens, testing how AI can apply reasoning to industrial process control like a human operator.
Github: https://github.com/andreenko-repo/ai_process_operator
A small web app that uses local Qdrant database and a camera to identify equipment and display information about it.
Github: https://github.com/andreenko-repo/shazam_for_plant_eqp
A proof-of-concept application demonstrating the use of a Large Language Model (in this case, Google's Gemini 2.5 Pro) to perform question-answering on Piping and Instrumentation Diagrams (P&IDs).
Github: https://github.com/andreenko-repo/pid_diagrams_llm/
A demo can be found here: https://huggingface.co/spaces/AntonAndreenko/pid_diagram_qa
A sample application that demonstrates how to use the Google Gemini API's native Google Search tool for two distinct purposes: data collection and response grounding. These techniques allow you to minimize hallucinations in many real-world applications.
Github: https://github.com/andreenko-repo/gemini_market_analysis
I converted the alarms dataset into a text format to test alarm management methods using NLP AI models.
The alarm viewer can be found here: https://huggingface.co/spaces/AntonAndreenko/tep_alarms_viewer
Github: https://github.com/andreenko-repo/tep_alarms_text
This is a small experiment in PID performance evaluation and tuning using historical data.
A demo can be found here: https://huggingface.co/spaces/AntonAndreenko/pid_tuning