Store scanning robots will get AI, object recognition boost with recent acquisition

Bossa Nova's acquisition of HawXeye follows big funding rounds for the retail robotics company
Written by Greg Nichols, Contributing Writer

Bossa Nova, the company behind the shelf-scanning robots that Walmart has been experimenting with in select retail locations, just acquired artificial intelligence company HawXeye.

A Carnegie Mellon spinoff, HawXeye has a library of 52 patents around face detection and object recognition, increasingly important technologies as robots in a variety of fields begin working with diverse object sets, such as SKUs in a store or logistics facility.

As part of the deal, HawXeye founder Marios Savvides, a CMU professor, has joined Bossa Nova and will serve as Chief AI Scientist. He'll primarily be working in areas like deep learning, facial biometrics, and object recognition.

Savvides is charged with advancing Bossa Nova's product recognition capabilities at scale and identifying out-of-stock and misplaced products.

Bossa Nova's robots rove stores scanning shelves and alerting managers to any issues with inventory, including misplaced or low-stock items. As I've written, the robots are also Big Data mining machines writ small, able to track product performance practically in real-time.

That kind of feedback gives brick-and-mortar retailers some of the same kinds of data that Amazon has long used to predict customer behavior.

"Bossa Nova is changing the retail ecosystem. They deliver unparalleled real-time product data to the largest retailers in the world through fully autonomous service robots," said Prof. Marios Savvides, Chief AI Scientist of Bossa Nova and director of CyLab Biometrics Center at CMU. "When combined with their robotics technology, our AI can unlock unprecedented levels of accuracy, reliability, and scale in retail scene understanding. With Bossa Nova, we're confident we can create a new era of retail data and analytics."

Bossa Nova recently announced $29 million in additional funding, bringing the company's total company to date to $70 million.

Editorial standards