SnackBox
Snack consumption is rising while physical activity declines — but health tracking systems treat diet and activity as separate data silos, missing the relationship between the two.
Built an IoT snack box prototype with embedded sensors that logs snack events, integrates Apple Health activity data, and uses AI to generate personalized lifestyle reports connecting diet habits with physical activity.
Working proof-of-concept demonstrating end-to-end data flow from physical snack event capture through health data integration to AI-generated daily and weekly lifestyle reports.
Overview
Most health apps track either what you eat or how much you move — never both together. SnackBox bridges that gap by turning a physical snack container into a data collection device. Every time you reach for a snack, the box records it. That event data flows alongside your Apple Health activity metrics into an AI analysis pipeline that reveals patterns you wouldn't notice on your own: do you snack more on low-activity days? Does your evening snacking spike after skipping lunch? The system makes these invisible connections visible.
Concept
SnackBox is not a consumer product — it's a concept validation prototype exploring how IoT devices can connect diet behavior with physical activity data. The core hypothesis: if you can see the relationship between your snacking patterns and your daily movement, you'll naturally make better choices. The system treats the physical snack box as an input device for lifestyle data collection.
System Architecture
The data pipeline flows: SnackBox IoT device (sensor event capture) → Data transmission → Apple Health API (activity data pull) → Data processing layer (merge & clean) → Google Sheets (storage) → AI analysis engine (pattern detection) → Report generation (daily + weekly summaries). The architecture intentionally uses lightweight tools (Google Sheets, API integrations) to validate the concept before investing in production infrastructure.
Key Highlights
Diet + Activity Connection
bridges the gap between food consumption tracking and physical activity monitoring that most health systems treat separately
Physical-to-Digital Data
turns a mundane physical object (snack box) into a behavioral data collection device without requiring manual logging
Concept Validation Approach
prioritizes proving the data hypothesis over building production infrastructure, using lightweight tools for rapid iteration
Data Flow
Snack Event Capture
The physical SnackBox prototype contains embedded sensors that detect when a user opens the box and removes a snack. Each event is timestamped and logged — capturing the raw behavioral signal of snack consumption without requiring manual food logging or user input.
Activity Data Integration
User activity data is pulled from Apple Health: steps, active energy burned, height, weight, and other movement metrics. This provides the physical activity context needed to understand snacking behavior — turning isolated snack events into data points within a broader lifestyle picture.
Data Processing & Storage
Snack event data from the IoT device and activity data from Apple Health are merged in a processing pipeline and stored in Google Sheets as a lightweight data management layer. This approach prioritizes rapid prototyping over production-grade infrastructure — the goal is validating the data connection, not building a scalable backend.
AI Pattern Analysis
The combined dataset feeds into an AI analysis layer that identifies patterns across snack consumption and activity levels. The system detects correlations like increased snacking on low-activity days, time-of-day consumption patterns, and weekly trend shifts — insights that are invisible when diet and activity data live in separate apps.
Lifestyle Reporting
Analysis results are delivered as daily summaries and weekly reports, showing the user their snack-to-activity relationship over time. Reports visualize consumption patterns alongside movement data, making abstract health data tangible and actionable for behavior change.
Gallery
Technical Details
Limitations & Next Steps
Physical Design
The prototype hardware needs further industrial design refinement — current form factor is functional but not consumer-ready.
API Dependency & Latency
The system relies on multiple external APIs (Apple Health, Google Sheets) which introduces processing latency and points of failure that would need to be addressed for real-world use.
Key Learnings
Connecting disparate health data sources (diet + activity) reveals behavioral patterns that neither source shows alone
Lightweight prototyping tools (Google Sheets as database) enable faster concept validation than building full infrastructure upfront
Physical IoT devices as passive data collectors reduce user friction compared to manual food logging — but sensor reliability is critical
AI pattern analysis is only as good as the data pipeline feeding it — data quality and synchronization matter more than model sophistication