Nikolai Kummer, P.Eng. Email & Phone Number

Calgary, Alberta, Canada

• Reduced required inventory presentations by 15% through the implementation of machine learning optimization algorithms.• Increased warehouse storage density by 20% by developing optimization strategies for inventory stacking and placement, maximizing space utilization and reducing overhead.• Developed an automated column inspection system integrating advanced signal processing and machine vision, reducing per-unit inspection time by 90%. Eliminated the need for personnel to physically enter columns, enhancing safety and operational efficiency across multiple sites with numerous units.• Engineered outlier detection methods for analyzing robot data (text logs and measurements), enabling preemptive flagging for repairs through real-time analysis of drive measurement information, preventing potential issues before they occur.• Led the optimization and deployment of a machine learning system in collaboration with the Alberta Machine Intelligence Institute, serving as the resident Attabotics expert. My code optimizations made the system operational, significantly boosting our automated warehouse throughput. This work culminated in a co-authored paper presented at the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023)Technologies: Python, C#, KQL, SQL,NLP, scikit-learn, plotly, keras

• Conducted in-depth analysis of robot runtime data and system interactions, identifying key areas for improvement and inefficiencies. This analysis initiated new projects that significantly enhanced system performance.• Spearheaded the ETL data pipeline overhaul, significantly reducing Azure resource usage and cutting cloud expenses.• Developed data pipelines to proactively monitor data streams from a fleet of robots and automated systems, incorporating advanced IoT edge data collection, cloud analytics functions, and proactive notification functions to alert personnel of potential issues before failure occurs.• Developed a centralized message routing system utilizing Azure, IoT Hubs, EventHubs and Azure Data Explorer to collect logs and monitor robots in real-time for unparalleled insights and proactive maintenanceTechnologies: Python, C#, Kusto Query Langauge, Azure Functions/Event Hubs, IoT hubs/Storage/Stream Analytics/Azure Data Explorer, MLOps, Airflow

Calgary, Canada Area

• Engineered and optimized a smooth robot pick-arm controller for precise product bin placement on shelves, minimizing jarring and enhancing operational smoothness.• Created visualization Software using Python, Flask, and Plotly that empowered field personnel to analyze and visualize robot logs and detect potential issues.• Created automated robot log collection software that enabled fleet-wide robot monitoring.Technologies: Python, C#, Control Systems, Plotly, Flask

Calgary, Canada Area

o Expertise in crafting deep learning models, including convolutional neural networks (CNNs), for analyzing and interpreting fiberoptic distributed acoustic sensing (DAS) and distributed temperature sensing (DTS) data for regression and classification tasks.o Specialized in automating anomaly detection for well operations using diverse data sources, leveraging advanced modeling techniques to predict and identify anomalous behaviors.o Developed scalable, GPU-optimized full-stack analysis software with clustering, signal processing, and neural network capabilities to efficiently handle and process terabyte-scale datasets.o Skilled in designing user-friendly GUI applications in Matlab, Java, and Python, deployed to Azure which streamlined data analysis and empowered project managers and analysts to conduct faster, more efficient analyses.o Achieved a significant reduction in analysis time, cutting it down from weeks or months to just days or hours, enabling prompt reporting of downhole well issues to clients.

Rock Lake Provincial Park, Alberta, Canada

I managed the operations, safety, and well-being of 4 staff and up to 30 guests at a remote wilderness lodge in Rock Lake Provincial Park.o Responsible for guests and day-to-day operations at a remote wilderness hotel in Rock Lake Provincial Park.o Performed maintenance of critical systems (diesel generator, septic systems, water well), planned supply runs of food, fuel and other critical components twice a week. o Guided guests on hiking and fishing trips in areas that were outside of emergency services range.

Leduc

I sought out new R&D projects that would create new product lines or reduce manufacturing cost/increase tool life of existing products to make Ulterra more competitive in a low-oil price market. I used data science techniques to create new insights into the design process using unstructured historical data.Data Analyst:o Analyzed large bit run databases to identify market trends lading to the of design a patented mudmotor transmission, influencing future downhole tool sizing, using R.o The database entries were highly variable and involved extensive feature cleaning and automated text analysis.o I analyzed downhole data from the flowloop that I instrumented to provide the customer with data-supported evidence of our tool operation.Design Engineer:o I designed and instrumented a test stand that better quantified the effect of component wear on overall downhole tool function, which led to better maintenance practices.o I contributed component designs, planned experiments, and performed destructive tests on new prototypes.o Named inventor on a novel mud motor transmission patent.

Okanagan Campus

I introduced mechanical and electrical engineering students to control systems concepts and Matlab programming using a variety of media targeted towards diverse learners. Sessional Instructor (ENGR 315):o My supplementary explanation videos have now over 1,000,000 views on YouTube.Teaching Assistant:o I received the “Teaching Assistant and Tutor Excellence Award” for communicating complex concepts in an approachable manner.o I taught statistics for engineers and report writing/data visualizations as a teaching assistant.

Kelowna

o My research involved the fields of machine learning, control systems, and machine vision.o I created Adaboost.MRT, an ensemble multivariate regression algorithm that addresses issues with the Adaboost.RT algorithm and increases accuracy on multivariate regression problems.o In my thesis, entitled "Translational Visual Servoing Control of Quadrotor Helicopters", I implemented a vision-based quadrotor helicopter controller that achieved UAV positioning relative to a moving target.

Edmonton, Canada Area

o I designed and sourced components for an instrumented top drive test station that was successfully used to test and verify the operation of a new top drive design.o I designed parts and created engineering drawings (parts, weldment, and assembly drawings) for oil-field equipment in Autodesk Inventor under tight deadlines.o I closely worked with mechanical, electrical, and hydraulic engineers to design parts to code.

Edmonton, Canada Area

Engineering Student/Machine Shop WorkerI worked for 5 years part-time/full-time in the machine shop and became intimately familiar with all aspects of the manufacturing process and assembly of downhole mud motors.- I developed a Drilling Jar Placement Software for downhole tools, which has been used to optimize jar placement within the bottomhole assembly.

M. Sc. Mechanical Engineering, Control Systems, Robotics

B.Sc. Engineering, Mechanical