5 amazing new uses for AI in 2021

Artificial Intelligence (AI) has quickly emerged as a game-changing technology, often helping researchers uncover insights much faster than they otherwise could.

Here are five impressive ways that researchers have already capitalized on AI this year.

1. Finding Concentrated Areas of Localized Pollution

Duke University researchers developed an AI algorithm that relies on satellite imagery and weather data to show the city blocks that are worst hit by pollution. They were especially interested in pollutants measuring less than 2.5 micrometers in diameter since they can travel deep into people’s lungs and cause health problems.

The algorithm looks for hot and cool spots of air pollution with a 300-meter resolution. That’s approximately the length of a city block in Manhattan. This approach first examines weather data alone, then fine-tunes the findings with satellite imagery data.

The researchers centered their work on Delhi, India. However, they said the algorithm could easily teach itself to do the job in different locations or keep pace with how a city changes.

2. Locating Electric Car Charging Point Infrastructure Gaps

People are getting more interested in owning electric vehicles, but many still worry about finding places to charge those cars. Doing that is usually not as straightforward as finding a gas station.

However, researchers built an algorithm that can mine through the unstructured data of charging point reviews. It achieved a 91 percent accuracy rate after training. The team chose eight main topics and 32 subtopics that characterized users’ charging behaviors covered in reviews.

Some of them related to whether there were enough functioning charging stations in their communities. Others discussed the cost to charge a car and whether charging stations helped people connect with fellow electric vehicle owners.

The researchers believe their use of AI could help shape future decisions about charging infrastructure and show where shortcomings exist.

3. Increasing Radiologists’ Productivity During the Pandemic

The COVID-19 pandemic placed health care workers under tremendous strain. Many hospital wards filled up, and staff scrambled to provide top-quality care as more patients arrived and resources became scarce. Technology can help alleviate some burdens.

A recent study showed that AI could aid radiologists in staying more productive when analyzing chest CT scans during the pandemic. More specifically, it reduced report turnaround time by an average of 30 percent, equivalent to seven minutes per case.

AI solutions can also boost the return on investment for radiology practices that use them. For example, some analytics platforms identify patients that need future appointments, such as for prescribed imaging or the removal of implanted devices. Those can increase the overall scan volume for a facility, making AI pay off.

4. Linking Climate Change, Human Land Use, and Animal Genetic Diversity

Scientists show ongoing concern about how climate change and the ways that humans use land could negatively affect genetic diversity in the animal population. However, until now, they haven’t categorized the effects on a global scale. A group at the University of Copenhagen depended on AI to make a world map showing the possibilities.

They combined information from animal gene banks with possible scenarios, such as if global temperatures rose and people converted tropical forests to agricultural lands. AI lets them add geographic information to genetic samples and build relevant models.

This approach enabled them to see where and how changes would affect animals the most. People could use that new knowledge to make policy choices that shape animal and land conservation efforts.

5. Learning More About Ocean Litter

Scientists know that ocean trash is a major concern, but they also recognize that addressing the problem is a daunting task. A recently developed AI app called MARLIT provides more than 80 percent reliability in classifying and quantifying marine litter.

The application focuses on macrolitter rather than the tiny particles of trash also found in the ocean. Until now, the main methods of spotting the garbage required direct observation, such as from people in planes or boats.

However, the ocean’s vastness and the volume of litter in it made it challenging to maintain productivity with those techniques. This approach blends automated aerial photography with AI algorithms. The results showed it was an efficient, first-of-its-kind effort in monitoring ocean trash.

AI Gives Impressive Results

These five examples show that artificial intelligence can help people cover new ground. Whether that means people find more efficient ways to work or examine issues with fresh perspectives, AI should continue to stimulate meaningful scientific and societal advancements.

Image credit: Laurent T / Shutterstock

Devin Partida writes about AI, apps and technology at ReHack.com, where she is Editor-in-Chief