Microsoft EPIC Regional Winners

The Microsoft EPIC regional semi-finals have concluded, and the winning teams are preparing for the final pitch competition at the Microsoft Technology Center on Nov. 16th!

Follow all the action on Facebook – https://www.facebook.com/iamSTEMpowered/

The pilot launch of the Microsoft EPIC program in Houston included a series of intensive day-long events at each of the area Microsoft Retail Stores – Woodlands, Baybrook, and Galleria.  Students received a crash course in the fundamentals of Artificial Intelligence (AI) and Data Science, and then applied what they learned to formulate AI technology solutions to real-world problems.

Following a series of lectures from university faculty and Microsoft staff, the teams worked with undergraduate coaches from the NASA SUITS program to create a “Shark Tank” style pitch, which they presented to a panel of industry experts and STEM education specialists.

The winning teams will face off in the final round on Nov. 16th, and the overall winners will be featured on a special segment of Great Day Houston w/ Deborah Duncan on Nov. 21st.

Baybrook Mall Semi-finalists

At the inaugural Microsoft EPIC session at Baybrook Mall, the winning team from Pearland Junior High West proposed the EVA AI system, which uses advanced machine learning, complex sensors, and behavior recognition algorithms to assess an individual’s driving, establish a normative baseline, and scan for aberrant driving behavior (intoxication, medical conditions, distraction, emotional issues, etc.).

By recognizing what constitutes “good driving” for each individual, the EVA AI system would identify and potentially mitigate high-risk driving behaviors in advance of a catastrophic accident.

In addition to drunk driving and medical conditions, this system would allow for dynamic control over performance of the vehicle, effectively using customized driver presets to govern the responsiveness, acceleration, suspension, etc. of the vehicle, based on the skill of the driver. 

Galleria Mall Semi-finalists

A team of students from St. Thomas More School won the second round of Microsoft EPIC with Project Aqua 05 – a complex networked AI system that uses an autonomous aquatic drone to identify relative levels of pollution in the ocean, in order to mitigate current pollution levels and prevent future contamination.

These 7th/8th graders developed a concept for a submersible drone system that employs a strong AI and machine learning to map levels of micro-plastics and other pollutants in the ocean.

At the most basic level, the Project Aqua 05 AI uses an array of onboard sensors and cloud-based data from satellite images, weather patterns, and Thermohaline circulation maps (ocean currents) to identify areas of increased pollution and call in clean-up crews in critical situations; BUT it doesn’t stop there!

Using predictive analytics, the system traces and identifies the likely source/origin of the pollutants, and anticipates the movement, shift, and accumulation over time!

Woodlands Mall Semi-finalists

Teams from West Briar Middle School and four Spring ISD middle schools (Bailey, Bammel, Twin Creeks, and Wells) competed in the last regional semi-finals for Microsoft EPIC.  The combined team of students from Bammel, Bailey, and West Briar Middle Schools won the Woodlands Mall session with an AI concept that learns from the individual lifestyle, behaviors, activities, and stressors for kids, and adapts a personalized stress management and productivity plan for each child.

Less Stress with AI aims to use a strong artificial intelligence system to identify patterns of causality between the student’s daily activities, emotional health, and  academic performance.  The camera and microphone of the student’s smartphone provides data points on relative stress, anxiety, and happiness, and this information is aggregated with self-assessment and activity data provide through opt-in push notifications that reward engagement and responsiveness. 

The proprietary APP also syncs with a teacher’s digital lesson plan and grade book to provide dynamic tracking of the student’s academic performance.  A machine learning algorithm analyzes the data and looks for trends – e.g. The student engages in regular physical activity, spends time with friends, allocates 2hrs/night for studying, and this corresponds to a higher happiness index and better test performance for the week.