post

Tomorrow’s innovators: First class graduates from Global Innovation Exchange

Archisa Guharoy and other graduates watch a video
Archisa Guharoy and other graduates watch a video during the graduation ceremony.

“When I measure the value of education, I look at those ‘aha’ moments. I had a lot of those moments here. I’ve been really happy with the program.”

GIX trains students to think globally and ethically in an era of rapid change and increasingly shorter business and technology cycles, says Vikram Jandhyala, GIX’s co-executive director and vice president for innovation strategy at the University of Washington.

“How can we build a future set of innovators who will create products and services and be part of organizations that can navigate this changing landscape and lead innovation?” Jandhyala asks. “That is the premise. It’s not business as usual.”

Most of the 10 team projects of GIX’s inaugural class were sponsored by the institute’s industry partners, which include Microsoft, Boeing, T-Mobile, AT&T and Chinese technology company Baidu. Company leaders pitched loosely defined projects to GIX students, then mentored the teams as they developed their projects, leveraging the companies’ technologies.

That approach to learning makes GIX unique, says Ranveer Chandra, chief scientist for Microsoft Azure Global and an advisor on the chicken-monitoring project, named Cluck AI.

“The students are able to take the latest research from industry, build on top of it and show what can be achieved,” Chandra says.

Students give a presentation
Students (from left) Xu Yan, David “Davo” Franco, Padraic Casserly and Ibtasam “Ibi” Sharif present their Cluck AI solution, which uses machine learning to identify when farm chickens are in distress.

“This kind of an industry-academic collaboration, where industry’s not just handing off something to students but providing them with cutting-edge research and working closely with them, is something I haven’t seen before as part of a curriculum.”

The Cluck AI team worked with Microsoft engineers to use machine learning to identify when chickens on a farm are in distress — when a predator is nearby, for example, or when they are overheated. A microphone captures the animals’ sounds and sends them to Azure storage containers. When an anomaly is detected, the audio data is pushed to a dashboard, along with an image of the chicken in apparent distress. The farmer, who may not be onsite to monitor the livestock at all times, gets an email notification and can assess what’s wrong.

“It’s kind of like a baby monitor for poultry farmers,” says team member Padraic Casserly, 32.

The solution uses technology developed for Microsoft’s FarmBeats project, an artificial intelligence and Internet of Things platform that harnesses data to increase farm productivity and cut costs. The four-member GIX team launched the project after Chandra, then the principal researcher behind FarmBeats, suggested that audio data could have untapped potential in farming.

The team decided to focus on chickens, interviewing farmers and even buying two chickens of their own — named Margarita and Daisy — to record them. The students hope to develop the solution further and are looking into possible funding for a start-up.

Leave a Reply