Search

Google Health 5.11 improves automatic workout detection

Translated title. Original: Google Health 5.11 migliora il rilevamento automatico degli allenamenti

Google Health 5.11 expands automatic workout detection to over 100 activities, covering Fitbit Air and Pixel Watch devices. The update uses multiple data signals and user feedback to improve accuracy and reduce manual input, with notifications sent ~30 minutes after activity and broader Android rollout ongoing.

GitHub - tdeemon/tdeemon: Self-hosted, AI-first nutrition tracking with an adaptive TDEE · GitHub

tdeemon is a self-hosted, AI-assisted nutrition tracker that measures real TDEE, logs meals, and proposes plan changes via an AI coach that the user must approve. It integrates with Apple Health and Open Food Facts, uses a SQLite database, supports Docker/Compose deployment, and offers data export and multi-user support.

GitHub - machinesr/Workout-Trainer: Study2Challange Hackathon 2025 · GitHub

This Python webcam app uses MediaPipe and OpenCV to track exercise form, movement timing, and repetitions across five exercises. The repository describes Version 1 and links to PermaFit v2 as a more user-friendly release.

Strava now connects with Google Gemini – artificial intelligence analyzes your training history - Muropaketti.com

Translated title. Original: Strava yhdistyy nyt Google Geminiin – tekoäly analysoi treenihistoriasi - Muropaketti.com

Strava now lets paid users link their training data to Google Gemini for personalized analyses, available in the US. Gemini can assess pace at a given heart rate, strength training impact, and recent race results against personal bests, with optional read-only access and easy opt-out.

Steps: Apple Watch Series 12 and Fitbit Edge compared | NeoTeo

Translated title. Original: Pasos: Apple Watch Series 12 y Fitbit Edge comparados | NeoTeo

NeoTeo reports a comparison by Jane McGuire from Tom’s Guide showing Apple Watch Series 12 registering 8,293 steps and Fitbit Edge 8,199 vs manual 8,270 during a three-segment walk, with Apple Watch closer to manual in all segments and a total deviation of about 0.28% for Apple Watch and 0.86% for Fitbit; Tom’s Guide notes a 50-step difference in published totals, and previous 6,000-step test favored Fitbit Edge. The article includes publication details and context about measurement methods.

I tested the Apple Watch's new treadmill credentials to see if it's really accurate at tracking distance without GPS

The author tests watchOS 27's indoor distance tracking across several Apple Watch models, using a Coros Pod 2 as a reference, and finds overall improvement with small remaining discrepancies, especially during intervals, suggesting indoor tracking is now more reliable though not perfect.

[For multi-store headquarters in retail and food service] Start accepting consultations for short workouts that fit the burden of standing work within store shifts - ASCII STARTUP

Translated title. Original: 【小売・飲食の多店舗本部向け】店舗のシフトの中でできる、立ち仕事の負担に合わせた短時間運動の相談受付を開始 : - ASCII STARTUP

A corporate health program by Fractal Workout offers 5-minute online exercises for staff in multi-store retail and food service, tailored to shift patterns. It includes monthly health monitoring, optional AI posture analysis, e-learning, and seminars, with a scalable deployment model and required store-site information for setup.

Life N'Joy focuses heavily on AI-powered training: 17 minutes are enough - Siljan News

Translated title. Original: Life N'Joy satsar stort på AI-styrd träning: 17 minuter räcker - Siljan News

A Swedish gym introduces seven AI-powered machines for a 17-minute, personalized circuit training; an initial check-up tailors machine settings to each user, making workouts efficient and engaging while promising measurable progress.

Fitbit Air vs. Whoop 5 am iPhone in the test: measure, interpret, coach | heise online

Translated title. Original: Fitbit Air vs. Whoop 5 am iPhone im Test: Messen, deuten, coachen | heise online

Two lightweight, display-free fitness bands, Fitbit Air and Whoop 5, compare in measuring daily activity, heart rate, sleep, and recovery, with different pricing tiers and features; the test finds similar goals but distinct approaches, and full analysis requires a subscription.

GitHub - devladpopov/personal-coach-kit: Your own AI personal trainer and nutritionist in 5 minutes: questionnaire -> program, meal plan from your store, tracker with timers, optional Telegram bot. Claude, ChatGPT, Gemini or local model. · GitHub

A GitHub project provides an AI-driven personal trainer and nutritionist toolkit. It uses a questionnaire to generate a customized 7-day training plan and meal plan, tracks progress, and optionally uses a Telegram bot. The kit runs with a single command, supports multiple AI providers, and stores data locally with options to export backups.