Smart Scale

To develop an intelligent, automated, and seamless system for reliable, AI-driven nutritional tracking, transforming manual logging into a "Scan and Track" paradigm.

Existing apps like MyFitnessPal and HealthifyMe rely on manual database search and self-reported portions, lacking precision weight sensing and vision-based food recognition.

Instant, verified carbohydrate counting for diabetic patients; granular protein tracking for athletes; general calorie awareness for wellness users.

Precisenutritionaltrackingisfoundationaltohealth, yetmanualandimage-onlymethodsare tediousandinaccurate. Thispaperintroduces Smart Scale, aconsumer Io Tdevicepairedwith Food Sense, a native mobile wellness platform that overcomes these challenges by integrating high-precisionweightsensingwithon-devicemultimodal AI. Smart Scalecombinesacalibrated HX 711 load cell on an ESP 32-WROOM with Food Sense’s dual-backend inference pipeline: (i) Google’s Gemma 4 E 2 B(Q 4 K M,3.2 GB) executedentirelyon-deviceviallama. cppwith Metal GPU acceleration, (ii) a YOLOv 8 TFL ite detectorfine-tunedon 94+ Indianfoodclasses, and(iii) anoptional Gemini Cloudproviderforuserswhosupplytheirownkey. Thesystemis shipped as a software under the Apache 2.0 license and ensures that food images never leave the user’s device by default. This work details the hardware, embedded, and on-device AI architecture anddemonstrates its potential toprovide reliable, private, subscription-freedietary monitoring.