From reducing childhood obesity trends to preventing falls with smart shoe insoles, a global contest has showcased big data’s powerful potential.
Analytics firm SAS’ first ever global Hackathon has attracted participants from more than 30 countries, comprising novices and experts alike, from its own customers and partners as well as startups and students.
Throughout March 2021, hundreds of curious minds, technology enthusiasts and data scientists and business visionaries were challenged in a host of industries to apply technology to solve real-world business and humanitarian problems.
Recently, regional winners for the Americas; the Asia Pacific region; and Europe, Middle East and Africa (EMEA) were announced. In addition to recognition, winning teams will earn an invitation to partner with SAS to commercialize their ideas. They will continue to enjoy access to the firm’s products as well as expert guidance, resources and support to continue development of their applications.
Applying data science in practice
According to the firm’s Senior VP of Customer Engagement and Support, Bob Messier: “The diverse teams in this global Hackathon sparked new ideas while addressing nearly 100 unique use cases. As a technologist, I’m amazed by the broad range of analytics technologies they used. All the teams deserve special recognition and thanks for their hard work and contributions.”
Some of the ideas and themes tackled include:
- Agricultural tech
The team NPK4Ever from North Carolina State University and startup Phinite created a model for manure-based fertilizer recycling that can improve the environment and food production. The model identified potential manure sources from large-scale animal farms across the US, depots where this fertilizer could be consolidated and dried; and factories to which it could be shipped for processing. The model also analyzed varying prices for carbon offsets.
- Environmental, social and governance (ESG)
The team from EMEA—KPMG Austria created a new data analysis feature for its ESG tool CLIMAID. The tool uses machine learning and natural language processing to analyze web data, score firms across 39 categories and build a model to predict ESG scores for any portfolio.
- Banking tech
Team ifb4Sustainability from Germany created a tool called Positive Impact Analyzer for Banking to help the financial sector explore its role in achieving the United Nations’ Sustainable Development Goals. The tool combines a Python model for calculating SDG index scores with a dashboard powered by visual analytics to investigate and explore data on bank portfolios’ past, present and future sustainability performance.
- Energy tech
Teameveris/Enel of Spain/Italy worked on a project to help an energy company develop predictive models that forecast external costs based on historical results. The team used visualized data and compared different models to determine the most accurate algorithm for long-term forecasting.
- Government/Public Sector tech
Team Hackanadians of Canada was assigned to reduce US firefighter deaths due to collisions with emergency vehicles. As many of these accidents occur at intersections, the team of data scientists created Traffic Lights for Life, an AI-based system that allows traffic lights to ‘listen’ for emergency vehicles through audio sensing, in-cloud deep learning and intersection control. The system protects first responders, who in turn protect society.
- Health care tech
Canada’s Red Hot Chili Steppers team from Pinnacle Solutions and the Ontario Tech University combined IoT sensors in a smart shoe insole with AI and machine learning tech to create a risk model that can help health care professionals analyze a patient’s movement and create a risk score for loss of balance and falling.
- Health/Childhood Obesity
Co-winners Tupã Fit [Brazil] and Digital Community Twin [Sweden] tackled the topic of childhood obesity. Through the ‘Brasil 2030’ project, researchers from São Paulo State University, NK Health Information System and Paraíba State University used digital technologies powered by AI to interpret data on childhood obesity and connect and educate people. This project is inspired by the work of researchers at Sweden’s Uppsala University, who used municipal and regional data to create a digital twin of Swedish society. Changes to the model allow researchers to study the impact of new programs on children’s health and move toward eradicating childhood obesity.
- Insurance tech
China’s Eagles team created a machine-learning model to explore and monitor big data on health insurance claims. The model serves as an early-warning system to detect abnormal behavioral patterns by doctors, including potential fraud.
- Manufacturing tech
Team LivNSense of India formed from a startup serving the manufacturing industry, created a solution combining industrial IoT and AI to optimize the performance (in terms of better energy use and lower emissions) for steam- and gas-fired furnaces. Its iSense41 technology transforms these furnaces into cognitive equipment that power smart factories
There was also Singapore’s team of students and professors from Nanyang Polytechnic, which developed an AI-powered system that applies pattern-recognition neural networks and machine learning to acoustic patterns for computer numerical control machines. Giving ‘ears’ to these common manufacturing machines, the data driven system can help monitor for and detect anomalies (such as tool wear or break) in real time, improving efficiency, predictive maintenance and worker safety, which are crucial elements in the manufacturing industry. With the proposed system, CNC machines will be equipped with an intelligent engine with the ability to have awareness of their real-time working conditions.
- Retail tech
Team 3KT of India worked on queue management by analyzing video data and training their model for optimizing the queue experience.
- Telecom tech
Team Telefónica España of Spain had a team of engineers used analytics to build, train and score models and generate forecasts, creating an updatable dashboard. With it, the team can monitor network performance, anticipate capacity and congestion problems before they occur, and estimate the resources needed to solve them. In this way, Telefónica can ensure the best possible service for its customers.
- Computer vision
Team ITsAmsterdam [Netherlands], comprising members from Amsterdam University Medical Center, Gelre Hospitals and ITsPeople, created a computer-vision model to help surgeons improve patient safety during cholecystectomies, and common but major surgeries to remove the gallbladder.
- Natural language processing
Team Butterfly Data of Canada used conversational AI to build a chatbot to deliver cancer lifestyle advice for patients and their families and friends. It provides statistical data from across Canada, nutritional information, directions to local health care services, and links to trustworthy websites.
Each team’s use case was evaluated on variety of criteria based on the Lean Canvas business model, including overall problem and solution, technology showcased, unique value, market potential, innovation level and business plan viability.
Said Einar Halvorsen, Global Hackathon Lead, SAS: “Along with solving big problems, this year’s global Hackathon was about creating the ultimate virtual learning and networking experience for refining existing skills and acquiring new ones. Gamifying the experience with some friendly competition, the Hackathon transcended participants’ backgrounds and geographies to foster collaboration and innovation that benefits everyone.”