�糵����

Developing a for construction workers netted the top prize at the SAS Institute’s annual for a team of Illinois Institute of Technology researchers, which is comprised of alumni and a current student.

StaSASticians earned the top spot in the hackathon’s Americas regional competition, as well as in the Insurance and Internet of Things categories. Brandon Sharp (M.S. AMAT 2nd Year), Narges Hosseinzadeh (M.S. AMAT ’23), Irina Klein (M.S. AI ’23), and their mentor Sou-Cheng Choi teamed with Irina Ivanova, a seasoned business analyst, and BeeInventor Ltd. The high-tech company brought in Jeffrey Li to join the Illinois Tech team. 

StaSASticians was among 145 teams—with 1,731 individual participants—from more than 70 countries participating in the 2024 SAS Hackathon. During the SAS Hackathon, teams of data scientists, business analysts, technology enthusiasts, and students come together—often virtually across continents—to network, learn from each other and from SAS mentors, and experiment with new technologies that spark innovation and improve lives.

Using artificial intelligence and SAS Viya Workbench, SAS Data Maker, SAS Viya Enterprise, and Python tools, the team constructed a heat stroke prevention system by collecting and analyzing data collected by construction workers’ helmets. SAS is an AI and data solutions provider.

“Unlike academic projects, which often focus primarily on model building, this experience highlighted that in real-world scenarios, most of the effort is dedicated to ensuring data accuracy, preprocessing, and structuring,” Sharp says. “This invaluable experience provided us with practical skills that are directly transferable to addressing real-world problems in future projects.”

The system utilizes real-time weather data and sensor information collected by BeeInventor, an Internet-of-Things innovator, via the company’s smart helmets. The helmet sensors measure workers’ physiological factors such as heart rate and core body temperature. Utilizing the AI models developed by StaSASticians, heat-stress illnesses could be accurately predicted and prevented with intervening measures.

The team used SAS’s open-source compatible and cloud computing environments for big data preparation and automated AI model development. In addition, SAS Data Maker was applied to generate synthetic data to address data scarcity.

“Our team spent a significant portion of time exploring and processing the raw datasets into normalized, clean dataframes ready to be used for modeling,” Klein says. “A lot depends on the decisions that are made early in the pipeline about data processing, filtering, and the treatment of outliers and null values.”

The team’s innovative solution is timely given the increasing temperatures in many cities and countries due to climate change. By implementing this system, StaSASticians’ aim to help construction companies, as well as insurance corporations, save the lives of construction workers, increase safety in risky environments, enhance worker productivity, establish regulatory compliance, and significantly reduce insurance and health care costs.

“The lack of heat stroke prediction in the current construction industry poses a significant threat to both the health and lives of workers,” Li says. Heat stroke has a mortality rate as high as 80 percent with delayed treatment, while early diagnosis and immediate cooling can reduce that rate to 10 percent. Affected workers also may experience long-term negative health consequences, impacting their overall well-being and productivity.

“BeeInventor has extensive experience with IoT products for the construction field,” Li says. “During the SAS Hackathon, we tested several models and reviewed numerous features, from which we identified top features for heat stroke predictions and developed accurate predictions toward heat stroke risk five to 10 minutes in advance.”

The team was organized and mentored by Sou-Cheng Choi, research associate professor of applied mathematics at Illinois Tech. She connected the alumni and the student with BeeInventor and put them on the path to developing the heat stroke prevention system.

Choi says the SAS Hackathon gives students and young professionals hands-on experience by providing them with an opportunity to apply theoretical knowledge to solving pressing real-world problems.

“Our team’s working and winning experiences in the 2024 SAS Hackathon have not only enhanced our technical knowledge and capabilities with SAS Viya Enterprise and SAS Workbench, but also fostered personal development in communication, collaboration, and confidence,” Choi says. “These events also provide invaluable networking opportunities with industry professionals and building a portfolio of projects that showcase their abilities in their resumes.”

Additionally, the CEO of BeeInventor, Harry Chan says, “We plan to conduct a construction field study next year to gather high quality data and build a highly accurate heat stroke prediction system using the team’s analysis and models as a basis. Academia-industry collaboration will play a critical role in achieving this goal, and we are enthusiastic about exploring the potential of such partnerships with Illinois Tech.”

The overall SAS Hackathon winner will be named at SAS Innovate, the SAS data and AI technology conference, May 6–9, 2025, in Orlando, Florida. StaSASticians and other winning teams are invited to share their projects in the conference and, as an additional recognition of their excellence, will be eligible to compete for the coveted Global Championship title of the SAS Hackathon.

Image credit: StaSASticians. The team photo was taken on the SAS Hackathon deadline on October 18, 2024. From left to right, first row: Sou-Cheng Choi, Narges Hosseinzadeh, and Jeffrey Li. From left to right, second row: Brandon Sharp, Irina Ivanova, and Irina Klein.