IoT Pet Collar with Geofence and Activity Tracking

Pets slip out of gates and go missing every day, and owners discover it hours later when someone notices the empty yard. Commercial GPS pet trackers exist but charge subscriptions and hide their workings. This project builds an open pet collar on an ESP32: a NEO-6M GPS module for position, an MPU6050 IMU that classifies rest versus activity, and firmware that defines a home geofence — crossing it raises an immediate alert on the owner's phone dashboard, which also shows the track history and a daily activity summary. The collar is a genuine wearable design problem: weight, battery life and a pet-safe enclosure all constrain the build, and the report documents those trade-offs honestly. GPS accuracy is the NEO-6M's rated outdoor performance, stated as a datasheet value, not a measured claim.

1. The GPS, IMU and ESP32 are assembled in the lightweight enclosure and fitted to a standard collar strap.
2. The owner sets the home coordinates and geofence radius in the firmware configuration.
3. On power-up the collar acquires a GPS fix (outdoors) and starts publishing position and activity over Wi-Fi.
4. The IMU continuously classifies rest versus movement; the firmware accumulates active minutes.
5. Crossing the geofence boundary triggers an immediate alert on the dashboard; the track history records the excursion.
6. The dashboard renders the live position, track history and daily activity summary for the owner.

Tech stack:

• ESP32 development board (Wi-Fi)
• NEO-6M GPS module
• MPU6050 IMU
• Li-ion battery + charging module
• 3D-printed pet-safe enclosure
• Blynk / ThingSpeak class dashboard
• Status LED
• Arduino IDE (C/C++ firmware)

• [GPS geofence alerts] A configurable home radius around the house; the collar raises an immediate dashboard alert when the pet crosses the boundary.
• [Live location tracking] GPS fixes are published over Wi-Fi to the dashboard with a track history of where the pet has been.
• [IMU activity classification] An MPU6050 distinguishes rest from movement, building a daily activity summary (active minutes vs rest).
• [Daily activity summary] The dashboard aggregates active time per day so owners can spot lethargy trends worth a vet visit.
• [Low-battery alert] Battery voltage is monitored and a low-battery alert is raised before the collar goes dark.
• [Pet-safe wearable enclosure] A lightweight 3D-printed enclosure with rounded edges and a standard collar strap mount, documented with a weight budget.
• [On-collar status LED] A small LED signals GPS fix and alert states without needing the phone.

• ESP32 collar electronics with GPS and IMU in a pet-safe enclosure
• Collar strap mount and weight-budget documentation
• ESP32 firmware (GPS, geofence logic, activity classification, dashboard)
• Enclosure 3D-print files and assembly guide
• Project report PDF, PPT presentation and viva Q&A document

• GPS needs open sky — indoors or under dense cover, fixes degrade or drop; the dashboard shows last-known position honestly.
• Tracking range is limited to Wi-Fi coverage; this is a home-and-neighbourhood tracker, not a cellular one (cellular variant is future scope).
• ±2.5 m GPS accuracy is the module's datasheet rating outdoors, not a measured claim.
• Battery life depends on the GPS fix rate; continuous tracking trades against the design target of a full day.
• Activity classification is rest-vs-active only — not a veterinary diagnostic.