Projects Prove the Power of Data Analytics
MIT student teams deliver solutions to business partners
By Paula Klein
The power of data analytics to solve — and to head off — real-world problems was proven again at MIT this month.
At the December 2 MIT Analytics Lab (A-Lab) presentation day it was clear that the “lab” work in this case, is done in the real world. Twenty-one graduate student teams described the results of their semester-long projects — from improving financial trading data and sales forecasts, to optimizing call center systems, and fine-tuning weather predictions and land mind detection. Their sponsors can be found among the Fortune 100, leading startups, and in the public and non-profit spheres including Netflix, Visa, Mass Mutual, AB InBev, and Schneider Electric.
The partnerships demonstrate the practical value of data science in today’s business world. Three winning teams were announced at the in-person event attended by about 120 people, and the traditional MIT trophy was passed on to the Handle Global team. Handle Global is a privately held U. S.-based healthcare supply chain and fulfillment company. It uses analytics and AI technology to deliver proprietary platforms and systems to its medical customers.
The Sloan graduate students — Austin Ader, Cameron Cubra, Evan Marrone, and Estella Dentinger (pictured above) — showed “thoughtfulness and creativity” in their end-to-end solution, and in development of a dashboard and an API,” according to, Tod Loofbourrow, CEO and Chairman, ViralGains, who served as one of the judges.
The sophisticated asset management model carried out Handle’s mission of developing “technology solutions that transform the healthcare capital equipment supply chain.” In 2021, the company was included in the Inc. 5000 list which ranks the nation’s fastest-growing private companies.
Rx for Healthcare Efficiencies
The winning A-Lab project, Predicting Date of Device Manufacture When Purchase Date is Unknown, was narrowly focused yet the results will have wide impact for the company and possibly the healthcare industry overall, according to the company.
To find the optimal model of products and purchase dates, the team had to cull through 1.5 million different data assets.
Details of all of the A-Lab projects are proprietary, but Handle said that identifying product dates for thousands of assets makes it easier to know when equipment needs to be replaced or updated, and will save the company substantial money as well as time.
A-Lab students work closely with their sponsor organizations and MIT mentors to grasp the nuances of the business. Now in its ninth year, the seminar course taught by MIT IDE Director, Sinan Aral and Abdullah Almaatouq, requires student teams to select and deliver a project using analytics, machine learning, and other methods of experimentation to “develop results that diagnose, enable, or uncover solutions to real business issues and opportunities.”
Also judging this year were Yael Avidan, Senior Director, Product-Core Shopping Experience, at Chewy, and Renée Richardson Gosline, Senior Lecturer and Research Scientist, MIT Sloan School of Management and a MIT IDE Research Lead. The four key criteria they weighed are: Technical and analytical advances; effort, contribution, and how much improvement was delivered; business impact, and presentation.
Identifying Land Mines
The second place team represented the U.N. Mine Action Service (UNMAS), an entity within the United Nations’ Department of Peacekeeping Operations that works to eliminate the threat posed by mines and other explosives. The UNOPS team studied U.N. Land Release in Colombia.
Clearing explosives paves the way for economic development, but measuring this impact has proved a challenge for the organization. The students were tasked with determining the effectiveness of the program using machine-learning models.
Their work “really hit the mark on AI” implementation, said Gosline, and it demonstrated how AI can be actionable and ethical. The students also “thought outside the box on technology use,” she said, by digitizing satellite images before and after mine clearance to determine the effectiveness of the U.N. program. The project required knowledge of buildings, roads, and land development as well as the technical expertise. (Read more about this project here.)
Avidan said the third-place team, Schneider Electric, had a “really interesting and difficult problem to address, and student worked thoughtfully and considered the human effects as well as the machines.” Their topic, Control and Clustering for Smart Alarm Management, led them to work with AI and human operators to identify and classify the most severe and mission-critical calls among hundreds of messages coming into the data centers daily. Judges said the student presenters clearly explained their call reclassification approach where the most critical incident reports are now addressed immediately.
Loofbourrow noted that overall, “the quality and rigor of the A-Lab student work gets better and deeper into the businesses every year — and it’s reflected in their results.”
