AI drilling optimization: maximize ROP & extend tool life with continual learning - ON-1183

Project type: Innovation
Desired discipline(s): Engineering - computer / electrical, Engineering, Computer science, Mathematical Sciences
Company: I2G SOLUTIONS INC.
Project Length: Longer than 1 year
Preferred start date: As soon as possible.
Language requirement: English
Location(s): ON, Canada
No. of positions: 2
Desired education level: Master'sPhDPostdoctoral fellow
Open to applicants registered at an institution outside of Canada: No

About the company: 

i2G Solutions Inc. is a Canadian technology company delivering innovative AI-driven data solutions on Open Subsurface Data Universal (OSDU) DataPlatform for the oil and gas and broader energy sector through its flagship software, i2G SEDD (Subsurface Energy Data Discovery). I2G SEDD represents a next-generation data solutions to energy companies, is OSDU native, providing Machine Learning Toolkits and Agentic AI framework, to fundamentally modernize how subsurface and operational data are managed, discovered and utilized.
At its core, SEDD addresses long-standing industry challenges related to fragmented legacy systems, limited interoperability, and labor-intensive workflows, issues that disproportionately affect small and mid-sized energy operators. By leveraging the open, cloud-native OSDU standard developed under The Open Group, SEDD enables vendor-neutral data management, improved transparency, and a scalable digital foundation for responsible resource development.

Describe the project.: 

This research project aims to develop an adaptive AI-driven drilling analytics system that improves drilling performance, safety and operational efficiency by leveraging machine learning models on real-time drilling data.
Drilling operations generate large volumes of multivariate time-series data, including parameters such as weight-on-bit (WOB), rotary speed (RPM), torque, standpipe pressure, flow rate, and rate of penetration (ROP). These signals reflect the complex interaction between drilling equipment, operational parameters, and geological conditions. Recent research shows that machine learning models can analyze such telemetry data to detect abnormal drilling events and support operational decision-making in real timeThe main goal of this project is to design an AI-driven continuous analytics framework that delivers two core capabilities in a real-time system:
1. Real-time abnormal condition detection (early warning):
Identifies anomalies like stick-slip, kick, vibration, pack-off, and ballooning using Deep Learning for Time-Series Patterns (LSTM/GRU/CNN-LSTM/Hybrid with online learning, adversarial transformer architecture, Adversarial Domain Generalization on LSTMs,..). It provides low-latency alerts with minimal false alarms and leads time for intervention
2. AI-driven drilling optimization: focus on predicting drilling performance and recommending optimal operational parameters. Potential techniques include ROP prediction models: Knowledge-Guided Temporal Graph Neural Network, Hybrid Transformer LSTM,PSO-BP neural networks,… combined with Genetic algorithm or continual reinforcement learning to recommend drilling parameters under operational constraints.

A key objective of the research is to build an interpretable and generalizable modeling framework that supports continuous learning from accumulated drilling data while operating on real-time data streaming pipelines. The system will continuously improve as new drilling datasets become available and adapt to different geological environments.

The intern will work on developing Machine Learning workflows of above direction on Canadian, South East Asia and North Sea datasets.

The outcome will be a functional prototype to leverage AI/ML power to enhance Exploration & Production efficiency for Canadian oil & gas SMEs.

Required expertise/skills: 

The ideal candidate will have a strong background in Artificial Intelligence, Machine Learning, Data Engineering with hands-on experience in Python, data pipelines, APIs, and cloud-native systems. Familiarity with agent-based systems, workflow orchestration, and human-in-the-loop AI is considered an asset.
Equally important is domain knowledge in petroleum geoscience, drilling. Candidates with experience working with machine learning for well, production data will be well suited to the project.
Experience or interest in OSDU data platform, data quality control, metadata management, and interoperability frameworks is highly desirable. Strong analytical skills, problem-solving ability, and the capacity to collaborate across multidisciplinary teams—combining AI, data engineering, and petroleum geosciences expertise—are essential. Prior exposure to regulatory reporting, or energy-sector applications will be considered an advantage.