I am a systems engineer focused on developing new value chains within manufacturing and supply chain through the adoption of Industry 4.0 technologies such as the industrial internet of things and artificial intelligence with edge and cloud computing. In 2021, I started the Exponential Industry blog and newsletter to foster a community of engineers, managers, and observers as the Fourth Industrial Revolution takes hold.
I have delivered Industry 4.0 solutions across many manufacturing verticals including both fully automated and semi-automated manufacture and assembly. The specific processes that I have affected include: Magnetic resonance imaging (MRI) sensor manufacture, automotive powertrain assembly, high specification pressure module assembly, semi-automated apparel assembly, surface mount technology (SMT) line for computer chip assembly, aircraft final assembly, diesel engine repair, contact lens manufacture and package, automotive welding robot predictive maintenance, high-speed bottling operations, film manufacturing, and paper making.
Projects & Presentations
Mobile Handwritten Digit Classifier
A proof-of-concept application that leverages Flutter and TensorFlow Lite for classifying handwritten digits based upon the MNIST data set. The MNIST data set is the ‘hello world’ project for all machine learning practitioners. This application demonstrates how to deploy a trained neural network model onto a mobile device (Android and iOS) for use in a real-world setting.
Delivering Data Products for Operations
PyData Ann Arbor
Ann Arbor, MI
In this talk, David Rogers from Sight Machine will discuss the technical and non-technical aspects of delivering a scalable data product for use in enterprise operations. The aspects, ranging from data pipelining to customer education, will be blended with examples and anecdotes from my experiences delivering data products for some of the largest companies in the world.
AI for Manufacturing: Today and Tomorrow
2017 O'Reilly Artificial Intelligence Conference East
New York, NY
2017 O'Reilly Artificial Intelligence Conference West
San Francisco, CA
Artificial intelligence (AI) in manufacturing has been around for a long time, but are you aware of how it can make your operations more efficient and profitable? David Rogers explains how existing technologies like the digital twin approach, advanced decision making, and downtime cause detection have primed manufacturing for a profitable and efficient future.
A digital twin mirrors the entire production process, including machines, lines, and plants, and serves as the foundational layer for enabling advanced analytics and decision making on the factory floor. By adopting a digital twin approach, manufacturers can gather plant floor data and improve operations. With the digital twin in place, the digital thread will become a reality. Manufacturers today are using data to make decisions to improve the manufacturing process, but this can be complicated. AI will make it less so, as systems become more integrated and automated, providing manufacturers with an integrated view (instead of a siloed view) of the asset’s data throughout its lifecycle.
In addition, dynamic tolerance will be simplified. AI will solve one of the fundamental challenges that assembly line manufacturers face, which is determining where the different tolerances in the assembly of the product are and how best to optimize them. At the same time, predictive maintenance will become mainstream. Manufacturers will be able to count on machines to tell them when the best time is to replace a part or when a machine will need maintenance—and the overall costs associated with it; essentially manufacturers will be able to optimize variables to maximize profitability and operational effectiveness beyond just uptime. All of this will require less human interaction. In the future, machines will automatically react to countless factors, such as variation in material supply, that could introduce variability into a process.
NBA 2-for-1 Analysis
End of quarter situations provide a unique opportunity for a team to create an advantage by manipulating game flow. Improving awareness in these situations may allow for a team to “steal” a few points or even a possession. This article attempts to understand how manipulating game flow to create a ‘two for one’ situation is advantageous. Play-by-play data from 15,310 regular season and playoff games is systematically analyzed for ‘two-for-one’ sequences resulting in a sample size of 4,455 opportunities. In 55.4% of the ‘two for ones,’ the initiating team came out ahead. Curiously, only 17.8% of the time the initiating team ended up behind. Thus, the remainder of the time, 26.9%, the ‘two for one’ is considered a tie because both teams scored equal amounts of points to end the quarter. Overall, in a whopping 82.0% of the ‘two for ones’ identified; the initiating team ended the segment either tied or ahead!
San Francisco, California
6/2022 - Present
All your data, analytics and AI on one platform. A data lakehouse unifies the best of data warehouses and data lakes in one simple platform to handle all your data, analytics and AI use cases. It’s built on an open and reliable data foundation that efficiently handles all data types and applies one common security and governance approach across all of your data and cloud platforms.
I work at Databricks supporting large manufacturing and high technology enterprises in their adoption of artificial intelligence. I design the technical architecture and relate the technical needs to key business outcomes.
Director of Supply Chain Technology
1/2021 - 6/2022
Home Ready Plants Delivered to Your Door. Backed by 5 generations of Grow-How™. Get bigger, better, leafier plants delivered from our greenhouse to your house. 🌿🏡
I work at Bloomscape designing supply chain technology of the future to support the unique aspects of the live-goods value chain. I am designing a supply chain to drive efficiencies while enabling exponential growth in the business.
Lead Associate, AI Solutions Architect
1/2019 - 1/2021
At Booz Allen, we empower people to change the world. We bring bold thinking and a desire to be the best in our work in consulting, analytics, digital solutions, engineering, and cyber, and with industries ranging from defense to health to energy to international development.
I worked within Booz Allen’s Troy, Michigan office bringing new technologies and approaches to a variety of clients within manufacturing and supply chain. I supported sustainment, maintenance, and modernization efforts for major defense programs while building out enterprise strategy for providing AI, machine learning, and digital solutions to our clients.
- Strategic AI leader as applied to supply chains and industrial operations including serving as a capability lead for predictive part demand and mobile predictive maintenance solutions.
- Manager and responsible for the development of a team of data scientist, software developers, and business analysts.
- Lead internal investment effort for the creation of patent-pending technology to assess the health of machinery and assets in remote environments through the novel use of artificial intelligence.
- Created a process for assessing the suitability of an enterprise data environment for forecasting future part needs.
- Supported numerous business case analyses for supply chain management, advanced technology integration, and enterprise facility sizing.
- Programming: Python, R
- Data Science: PANDAS, Jupyter, Plotly, RShiny, TensorFlow 2.x, NumPy, SciPy
- Cloud: Docker, AWS
- Databases: MongoDB, SQLite
- Developer Tools: git, vim, linux
Lead Data Scientist
Ann Arbor, MI
7/2016 - 1/2019
Sight Machine is used by Global 500 companies to make better, faster decisions about their manufacturing operations. Sight Machine’s analytics platform, purpose-built for discrete and process manufacturing, uses artificial intelligence, machine learning, and advanced analytics to help address critical challenges in quality and productivity throughout the enterprise.
During my time at Sight Machine, I architected, developed, and deployed a cloud software platform for big data analysis within manufacturing. This included finding systematic approaches for the identifying and valuing prospective data sources, acquiring those data sources from the factory floor into a cloud environment, modeling the data within manufacturing specific concepts (cycles, downtimes, parts, defects, batches), structuring business problems into machine learning problems, developing business cases around cloud plaform enablement, and quantifying the value creation over multiple time horizons. I supported clients on four continents with a focus on both Europe and North America.
- Architected and led multiple large cloud PaaS deployments for global manufacturers such as Nissan Motor Co., Heineken N.V., and Westrock Co., growing the accounts from initial sale to proof-of-concept to enterprise expansion.
- Led a team of data scientists to create an SDK for scalable interaction with Digital Twins.
- Developed a number of novel machine learning approaches for understanding complex manufacturing phenomenon.
- Presented and communicated machine learning models and business cases for executives, engineers, and machine operators and built out account strategy.
- Assisted in new product development for a data analysis application suite tailored for manufacturing engineers, managed usage metrics, designed incremental features.
- Directly impacted the growth of major customer accounts enabling a successful $29MM Series C raise.
- Programming: Python, R
- Data Science: PANDAS, Jupyter, Plotly, TensorFlow 1.x, NumPy, SciPy
- Cloud: Docker, Kubernetes, GCP, AWS, Azure
- Databases: PostgreSQL, MongoDB, Microsoft SQL Server
- Developer Tools: git, kubectl, vim, linux, pylint, flake8, Jenkins CI, JIRA, Confluence
System Design and Analysis Engineer I & II
3/2014 - 7/2016
Boeing is the world's largest aerospace company and leading manufacturer of commercial jetliners, defense, space and security systems, and service provider of aftermarket support. As America’s biggest manufacturing exporter, the company supports airlines and U.S. and allied government customers in more than 150 countries. Boeing products and tailored services include commercial and military aircraft, satellites, weapons, electronic and defense systems, launch systems, advanced information and communication systems, and performance-based logistics and training.
At Boeing, I started my career as a full-stack software developer incorporating unsupervised machine learning techniques into digital signal processing systems. I became involved in every component of the systems development life-cycle including the analysis, design, development, test, implementation, documentation, and evaluation of RF systems containing integrated hardware and software components. I helped lead an effort to incorporate agile software development practices into the engineering process to improve first-time quality of the core product platform. I supported both internal research efforts and legacy platforms so that the firm could maintain its competitive advantage within RF systems engineering. I also helped bring adoption of Python into my working group improving the team’s rapid prototyping ability while growing my skills in Java, C++, and MATLAB.
- Awarded Boeing Knowledge Network Award for development and sharing of technical documentation which improved first-time quality and developer efficiency.
- Directed development of a Python SDK for enabling advanced signal processing plugins.
- Led an effort to virtualize and dynamically configure software environments in order to automate building and testing of numerous software components on a nightly basis using Vagrant and Puppet.
- Redesigned a software defined radio system improving system resource utilization and enabling production of a smaller form factor system to meet changing customer needs.
- Programming: C++, Java, Python, MATLAB
- Digital Signal Processing: FFT, Kalman filtering
- Cloud: Docker, Vagrant, Puppet
- Developer Tools: git, vim, linux, Jenkins CI, JIRA, Confluence
University of Virginia
MEng Systems Engineering, Accelerated Master's Program
2015 - 2016
The Accelerated Master’s Degree Program in Systems Engineering is a hands-on, practical program of study created by senior faculty at the University of Virginia’s School of Engineering and Applied Science in collaboration with the Darden Graduate School of Business Administration. It couples courses in business strategy and functionality with mathematics and modeling. The Program’s dual emphasis on engineering and business skills places graduates in a unique position to further any organizations' strategic goals.
My time at the University of Virginia formalized my knowledge in systems enginering while also developing my business acumen. The program’s emphasis on applied techniques in probability, optimization, statistics, and human factors left me with a toolkit for solving many problems facing leading organizations today. Additionally, business administration courses on topics such as business strategy, branding, marketing, negotiation, and risk analysis continue to serve me well as I help organizations find and capture areas for improvement.
Michigan State University
BSc Computer Engineering, Honors College
2010 - 2013
The Honors College at Michigan State University – one of the country’s most distinctive and extensive honors programs – serves academically talented, committed students who wish to pursue and achieve academic excellence. The Honors College strives to ensure an enriched academic and social experience for its members and to create an environment that fosters active and innovative learning.
At Michigan State University I pursued my passion for problem-solving through the rigorous study of computer engineering. Within my core curriculum I developed technical skills for designing and developing embedded systems including programming, software design, and circuit design. After class, I challenged myself further by assisting in undergraduate research in the Digital Evolution Lab under Dr. Ofria where I studied generic algorithms and artificial intelligence systems. Beyond the classroom I participated in student organizations such as the Association for Computing Machinery, Eta Kappa Nu (President 2012), and the Izzone.