Amazon wants users to get moving.
Amazon Halo, the company’s first fitness tracker and health platform, is introducing a new feature that will use smartphone cameras, artificial intelligence, computer vision and machine learning to recommend exercise routines like a personal trainer would, the company announced in a blog post Thursday.
The new feature, Movement Health, aims to create customized routines to help improve users’ stability, mobility and posture. It is slated to launch in the coming weeks.
Users will be prompted to take a five to 10–minute video assessment of themselves doing five movements: single-leg balances, forward lunges, overhead squats, overhead reaches and squats. Amazon Halo’s computer vision and machine learning tech will then evaluate the assessment with an overall “Movement” score out of 100, and users will get insight on their body’s mobility broken down into percentages, like hip and shoulder stability.
“This assessment delivers comparable accuracy to an in-person assessment with a professional trainer,” Amazon Halo’s senior technical product manager, Njenga Kariuki, wrote in the blog post, explaining that the algorithms assessing movement were made with a diverse range of body types.
The company also addressed privacy concerns in the blog post, saying that there are “multiple layers of privacy and security” built into the feature to protect data. Movement assessment videos, Amazon said, are deleted automatically within seconds of being processed.
Amazon Halo fitness band sparked backlash when it rolled out last year for having a body fat scanning and tone analysis feature. Critics like Gizmodo said it was invasive.
The new feature comes almost a year after Amazon Halo debuted as a health subscription service competing with the likes of Apple Fitness Plus, which also has content that tailors workouts to specific body types.