We at Inform believe that data tells a story, across all industries, and every week we’ll be rounding up the most interesting ones right here. This week: big data for speedier racers; fighting obesity; and helping schools help troubled students.
Formula 1 race cars are more than just pricey, speed machines. Nowadays, they’re giant data sensors, collecting factors such as the effect of stress and downward force on a car, brake temperature, tire pressure, and of course speed, and feeding that data back to analysts and engineers to measure how well the vehicles are performing.
The data is also modeled upon to obtain predictive intelligence on how the cars will perform in the future. In addition, vehicles are built for each track “based on historical data and simulations generated by the current season’s sensor data.”
However, data analytics aren’t the answer to everything. For example, it’s still impossible to capture “an accurate sense of where the cars are laterally on the track,” as well as “how well a tire is gripping the roadway.” In those cases, the best sensor is still the human driver.
Computer scientist and former army intelligence officer Paulo Shakarian is working on ways to use machine learning and big data to improve military intelligence.
While stationed in Iraq, Shakarian noticed that while intelligence workers like himself were tasked with analyzing all available data and hypothesizing possible courses of action, few actually had time for this in the midst of war.
This is where machine learning and big data come in. Shakarian’s research over the years has resulted in software used to detect IEDs, or improvised explosive devices, in Afghanistan; social media programs that help Chicago police fight gang activity; and a mathematical model of the behavior of ISIS.
The big data approach to medicine is quite different from health care’s traditional method of posing a hypothesis, devising an experiment, and testing. With big data, it’s about discovery — in other words, collecting huge amounts of data and seeing what patterns it returns.
The latter is the approach physicist Plamen Ivanov is taking at Massachusetts General Hospital in regards to the way organ systems interact. He and his team are collecting “hours and hours of data on vital signs,” such as that from EKGs, EEG, and ventilators, and seeing if they can “tease out how organ systems communicate with each other and coordinate behaviors.”
If his project is successful, Ivanov imagines a new kind of patient monitor that instead of just measuring blood pressure, heart rate, and brain activity, would “track the relationships between key organ systems — alerting doctors to cataclysmic phase changes in human health before they occur.”
A team at the University of Virginia’s School of Engineering and Applied Science is implementing a program that aims to capture data on environmental and behavioral factors that could contribute to childhood obesity.
Rather than relying self-reported or anecdotal data, the team is using an in-home monitoring system made up of sensors that monitor factors such as tone of voice, mealtime distractions, frequency of meals, and stress levels. The team aims to identify and model “preventable behavior patterns,” and if successful, plans to expand to other medical fields.
Sometimes just the existence of a massive amount of data isn’t enough. Tools are needed to streamline the collection and analysis process.
Some New York City school systems are facing this challenge and a nonprofit is helping by providing a way for the schools to feed the “vast supply of data” from multiple databases and sources into a single spreadsheet, and then training school officials on how to track student performance and devise plans to address issues and make improvements.
Before the implementation of this new system, some school workers had the arduous task of printing out reports and scouring them for patterns; using antiquated systems that resembled MS-DOS and required the typing of four-letter codes; and comparing paper documents and highlighting pertinent information by hand.
With the new tool, schools not only save time but are able to make decisions more efficiently and based on data rather than guesswork.