We apply the tools of big data, artificial intelligence, and problem solving in the following areas: Bioinformatics and Medicine, Climate Change, Energy, Environment and Ecosystem Management, Geographic Information Systems and Remote Sensing. Summaries of each active or past project are as follows.
Bioinformatics and Medicine
- Molecular Discovery: In Q4 2018 we developed a deep reinforcement learning system for finding similarly-structured molecules to a particular molecule. We then adapted this system to find molecules with high binding affinities to a particular molecule. Accuracy testing is still underway.
- Implant Design to Treat Type II Diabetes: We developed an implant design that holds promise as another step toward the treatment of type II diabetes. The details of this challenge are available here. We collaborated with a major pharmaceutical company and a research institution to implement and evaluate our design over the course of nearly a year.
- Real-Time Determination of Cardiac Spheroid Internal Pressures Using Photoelastic Polydimethylsiloxane Microspheres
- Carbon Sequestration: We are currently researching a sonochemical means of CO2 sequestration.
- Climate and Biome Model Interpretation: We have researched the migration of plants northward and upward due to climate change in the northern hemisphere, particularly in the Arctic.
- Novel, Efficient Wastewater Treatment: We conceived of a non-reverse-osmosis method to efficiently remove salt, boron, and organic materials from industrially produced water at scale. This water may be sourced from the ground through hydraulic fracturing, or produced in an industrial process. Nonetheless, the end goal is to return the water to a condition suitable for agricultural use. The details of this particular challenge are available here.
- Active Pumping of Oil through a Porous Hydrophobic Matrix to Achieve Rapid and Continuous Oil/Water Separation
- Cost-Effective Metering of Energy Use in an Ascending Pipes Distribution System
Environment and Ecosystem Management
AI-Driven Aerosol Monitoring System: In Q2 2018 we submitted plans for a system that would determine concentrations of particulate matter using input from an ultra-small NIR sensor. The system would use deep learning to compare sensor output with labeled data from an accurate sensor.
Geographic Information Systems
Geospatial Big Data Architecture and Engineering for Analytics: Beginning in Q4 2016 we began implementing a big data pipeline for a major biotech/agricultural services company operating in the St. Louis area. It incorporated a variety of geospatial data sources and was cloud-based.
Deep Learning for Image Analytics: In Q3 2017 we began planning and implementing deep learning software for an early-stage startup in the St. Louis area, Optar AI. The software was to run on a variety of remotely sensed image sources, to include drone imagery.
- Artificial Intelligence-Guided Navigation in Virtual Reality: We submitted plans for a system that would enable users to intelligently navigate a 3D environment to carry out mission objectives using AI-driven VR software. We also built out such a system and used it to navigate a virtual 3D environment.
- AI-Driven Moderation Helper: We developed an AI application to monitor Facebook groups, analyze content for problematic communication (profanity, hate speech, disruption), and to report that content to a public channel for moderators to investigate.
- Replacing Optical Glass with Transparent Zirconia