Here’s our November pick of the case studies inspiring us! Five amazing uses for data, analytics, ML and AI from across the globe.

1. “It would have recalled 34.1% of the women who went on to develop cancer”
UK-based Kheiron Medical Technologies is transforming cancer diagnostics by combining human intelligence with the power of AI. It won the UK Government’s AI in Health and Care Award for its astonishing product, MIA (Mammography Intelligent Assessment).

Trained on millions of images (“more than any radiologist would see in ten or even a hundred lifetimes,” says Kheiron’s chief strategy officer Sarah Kerruish), MIA has proven equal to a professional radiologist in detecting cancer on reading a mammogram. And, according to research by the University of Aberdeen, which included analyzing 220,000 mammograms from more than 55,000 people, in successfully identifying potentially missed cancers.

“Astoundingly,” says Dr. Miriam Stoppard, “the team found that MIA would have recalled 34.1% of the women who went on to develop cancer in between screenings.” And it does it all in seconds.

2. Like planting 37 million trees or taking 2.9 billion cars off the road
Vistra Corp. is a Fortune 275 energy company based in Texas. It operates power plants with the capacity to power nearly 20 million homes.

In 2020, Vistra built and deployed a heat rate optimizer (HRO) at its Martin Lake Power Plant. An AI partner worked with Vistra to optimize the HRO using a multilayered neural network. And when the models were accurate to 99% or higher, they were converted into an AI-powered engine generating operating recommendations every 30 minutes.

These recommendations helped Martin Lake improve efficiency by over 2% in just three months, resulting in savings of $4.5 million per year and abating 340,000 tons of carbon.

According to McKinsey, if this improvement was carried across all coal- and gas-fired plants in the U.S. electric-power generation industry, 15 million tons of carbon would be abated annually – the equivalent of decommissioning more than two large coal plants, planting 37 million trees, or taking 2.9 billion cars off the road – using only the data and equipment those plants already have. Read the full case study.

3. A 760% increase in sales
OlfinCar sells new and used cars in the Czech Republic. Already a good-sized company, with an annual turnover of around two billion CZK (around $89 million), it saw an opportunity to expand in a growing market. OlfinCar started by connecting data from three different sources: internal (orders, CRM, etc.), external (competitor and supplier data), and publicly-available (vehicle registration).

That data was centralized and cleaned, key systems were integrated, and work began. Competition monitoring made OlfinCar more competitive. Analyzing user behavior drove more effective and efficient advertising, while predictive algorithms more accurately identified customers’ needs at the website.

The results: OlfinCar reduced costs, increased turnover and – that headline number – increased sales by 760% in one quarter alone. Read the full case study.

4. Robot customer service agent saves lives
JD.com is the largest retailer in China. This year, it’s gone big on generative AI, upgrading its ChatJD robot customer service to provide shoppers with personalized recommendations and give sellers automated smart pricing and advertising tools. But that’s not all it can do.

JD’s customer service chatbot is, the company says, “able to detect the subtlest human emotions and interact appropriately”. Dr. Xiaodong He, deputy managing director of JD AI Research says: “Emotion is quite an important perspective of customer service. By implementing emotions in our robots, they can chat with customers in a more natural way.”

This has benefits beyond a good net promoter score that show the deep benefits of responsible AI by design.

A customer asked JD’s robot customer service how many sleeping pills would be required to take a life. The system quickly redirected the query to JD’s suicide hotline team, trained in psychological counseling, who took appropriate action. Since then, say JD, its chatbot, early intervention and quick response have saved tens of lives.

5. Counting elephants and identifying poachers
Two non-profits caught our eye this month, both wildlife conservation projects in Africa, both feats of ingenious engineering.

WildEye Conservation, which develops AI-based tools to facilitate the conservation of species, has developed TrapTagger, an open-source AI solution to count animals and classify species in images from camera traps. TrapTagger is being used by 40 organizations to process two million images a month, helping specialists to study the reintroduction of species into the wild.

Meanwhile, Strategic Protection Of Threatened Species (SPOTS) teamed up with FruitPunch AI to crowdsource a team of “AI for Good” engineers. Over thousands of free engineering hours, the team developed a poacher detection system and put it onto an autonomous drone.

According to their latest update: “The machine learning algorithms that spot the poachers have become more accurate, faster and more automated [but] the components are not quite ready for deployment in their edge computing entirety.” The team will keep going – for free – until they are.

Enjoying our articles? Sign up for our Weekly Insights newsletter to have fresh perspectives and thought leadership from Data Universe and our contributors delivered right to your inbox.