Final Project Proposal

ELI - Your Focus Training Desk Bot

ELI is an interactive desk robot designed to actively train and improve focus and productivity, rather than just tracking it. Unlike conventional focus apps that simply measure time spent on tasks instaed of the quality of the work done, ELI works as a personal productivity coach that learns your unique work patterns, corrects posture, optimizes your environment, and builds sustainable focus habits.

People nowadays struggle to focus at desks or any workspaces due to short attention spans, and poor setups. Good posture alone boosts focus by 20-30% through better brain blood flow, as shown in studies where standing desks improved executive function and working memory in high schoolers.I personally, face e hard time trying to just sit still and focus on my work to be productive.

So I decided to make ELI as my final project for this course. My personal desk bot that trains my focus instead of just tracking it.

Features of ELI:

Pattern Learning:Tracks when/where/how you focus best, for example "6 AM desk with gamma waves". Uses image processor and GPS to log conditions, then ELI suggests: "Your peak yesterday was: silence + upright posture yields better sessions."
Posture Correction:Camera detects slouching and vibrates a tethered "focus pebble" or says "Sit tall,good posture sharpens your brain 20-30%!" Backed by neurocognitive studies showing upright desks enhance executive function via frontal lobe activation.
Environment Optimization : Scans desk for clutter. For example,"Clear paper" or "declutter your desk" tidy spaces cut mental load by 15-20%.
Timers & Prosoche Training: Pomodoro or custom made timers. Starts with 5-min single task drills to enhance focus for example, "Breathe, focus only here, scaling up." Then after 90-min into work, ELI will be programmed to pause and display something like "Stand and stretch, good job!" To maintain a healthy working space.

How It Trains Productivity

ELI measures focus quality instead of the tasks you've finished,for example, posture compliance(upright vs slouch, high focus = head upright 90 % of the time) , over time logged, gaze stability. This builds attention habits, not dependency.

How It Works

Arducam Mega SPI camera captures real-time posture, gaze, and desk images for local processing.
A microcontroller like XIAO ESP32 C3 handles all logic: image analysis and pattern detection.
DS3231 RTC tracks precise time/date for peak focus correlation (e.g., 6 AM silence).
Push buttons (2x) enable navigation: select tasks, set timers, input weekly plans to SD.
Coin vibration motor buzzes whale body on slouch detection via camera thresholds.
OLED display shows eyes/expressions, timers, and prompts like "Sit tall, you're slouching!".
Speaker plays stored ringtones/voice cues from SD like "Breathe, focus here".
SD card stores everything: focus reports, patterns, ringtones and voice cues.
Offline focus trainer