BK-Shoot

Real-time Basketball Shot Tracker with Sensor Fusion

A low-cost IoT device that detects basketball makes/misses in real time using IR + vibration sensor fusion, paired with an Android app for live statistics tracking.

---

Demo

02_testing_video.mp4

---

Recognition

Award	Competition	Year
Honourable Mention	12th Planter de Sondeigs i Experiments (UPC, UAB, UB, Idescat)	2021
Presented	ICFO Young Photonics Congress	2021

"Per unir l'estadistica, el Big Data, la intel·ligencia artificial i la programacio a l'esport..." — Jury Citation (For combining statistics, Big Data, AI, and programming with sports)

View certificates and full details

---

Tech Stack

Layer	Technology
Hardware	Arduino Uno, IR Sensor (E18-D80NK), Vibration Sensor (SW-420), HC-05 Bluetooth
Firmware	C++ (Arduino), JSON serial protocol, sensor fusion algorithm
Mobile App	MIT App Inventor (Android), Bluetooth serial, SQLite storage
Data Analysis	SQL queries, statistical analysis by demographics

---

Architecture

graph TD
    subgraph HOOP ["BASKETBALL HOOP"]
        direction TB
        IR["IR Sensor<br>(E18-D80NK)"]
        VIB["Vibration Sensor<br>(SW-420)"]
        UNO["Arduino Uno"]
        HC05["HC-05 BT Module"]
        
        IR --> UNO
        VIB --> UNO
        UNO --> HC05
    end

    subgraph APP ["ANDROID APP"]
        direction TB
        BT_RX["Bluetooth Receiver"]
        STATS["Statistics Engine"]
        DB["SQLite Storage"]
        DASH["Live Dashboard:<br/>Attempts | Makes | Misses | FG% | Timer"]
        
        BT_RX --> STATS
        STATS --> DB
        STATS --> DASH
    end

    HC05 -.->|Bluetooth| BT_RX

Loading

---

How It Works

Sensor Fusion Algorithm:

The core innovation is combining two sensors to achieve accurate shot classification:

graph LR
    IR["IR Trigger<br>(ball passes rim)"] --> FUSION
    VIB["Vibration<br>(rim impact)"] --> FUSION
    FUSION["Fusion Window<br>(1000ms)"] --> CLASS["Classification"]

Loading

Scenario	IR	Vibration	Result
Clean swish	✓	✗	MAKE
Off-rim make	✓	✓	MAKE
Rim-out miss	✗	✓	MISS

Output Format (JSON over Bluetooth):

{"type":"make","ts":1234567}
{"type":"miss","ts":1237890}

---

Validation Results

Field-tested with 20+ participants and ~2,000 shots.

Metric	Result
Detection Accuracy	~95%
Mean FG% (players)	78.2%
Throughput	24.5 shots/min
Detection Latency	~250ms
Battery Life	>3 hours

Data Analysis

Shot data was exported to SQLite and analyzed by demographic factors:

Analysis by Dominant Hand	Analysis by Age Group	Baskets Distribution

See /testing/ for complete dataset and methodology.

---

Quick Start

# 1. Clone the repository
git clone https://github.com/omorros/bk-shoot.git

# 2. Upload firmware to Arduino
#    Open /firmware/bk-shoot.ino in Arduino IDE and upload

# 3. Build the Android app
#    Import /mobile/bk-shoot.aia into MIT App Inventor

# 4. Pair Bluetooth (HC-05, PIN: 1234)

# 5. Start shooting!

Pre-built APK: /mobile/bk-shoot.apk

---

Hardware

Bill of Materials (~€25 total):

Component	Model	Purpose
Microcontroller	Arduino Uno	Main controller
IR Sensor	E18-D80NK	Ball detection
Vibration Sensor	SW-420	Rim impact detection
Bluetooth	HC-05/HC-06	Wireless communication
Feedback	RGB LED + Buzzer	Audio/visual feedback
Power	5V USB Power Bank	Portable power

See /hardware/bom.md for complete BOM and /hardware/wiring-table.md for wiring details.

---

Challenges & Learnings

Building this project taught me:

Challenge	Solution	Learning
False positives from rim vibration without ball	Implemented sensor fusion with time windowing	Signal processing, debouncing
Bluetooth reliability issues	Added JSON framing and error handling	Serial protocols, data integrity
IR interference in sunlight	Tuned threshold and added shading	Sensor calibration, environmental factors
Data persistence in app	Implemented SQLite storage	Database design, SQL queries
Statistical validation	Collected data from 20+ users, analyzed demographics	Experimental design, data analysis

Technical skills developed:

• Embedded C++ programming
• Sensor fusion algorithms
• Bluetooth serial communication
• Mobile app development
• SQL and data analysis
• Scientific methodology (hypothesis, experiment, analysis)

---

Roadmap

v2.0 (Planned)

• React Native app (cross-platform iOS + Android)
• ESP32 with BLE (replace HC-05)
• Cloud sync with Firebase
• Web dashboard for coaches

v3.0 (Future)

• TinyML on-device shot classification
• Camera-based trajectory analysis
• Multi-hoop support for team training

---

Project Structure

bk-shoot/
├── firmware/           # Arduino C++ code
│   └── bk-shoot.ino
├── mobile/             # MIT App Inventor project
│   ├── bk-shoot.aia    # Source
│   └── bk-shoot.apk    # Pre-built APK
├── hardware/           # Schematics and BOM
│   ├── circuit_image.png
│   ├── bom.md
│   └── wiring-table.md
├── testing/            # Validation data and analysis
│   ├── 02_testing_video.mp4
│   └── *.png           # Analysis graphs
├── docs/               # Technical documentation
│   ├── tech-note.md
│   └── *.pdf           # Scientific poster & abstract
├── recognition/        # Awards and certificates
│   ├── AWARDS.md
│   └── *.pdf
└── README.md

---

Documentation

Document	Description
Technical Note	Methodology, results, and discussion
Scientific Poster	ICFO presentation poster
Abstract	Scientific abstract (English)
Awards	Competition recognition details

---

About

Author: Oriol Moros Vilaseca

This project was developed as a secondary school research project (Treball de Recerca) at age 15, combining my interests in basketball, electronics, and programming. It demonstrates end-to-end product development: from problem identification through hardware design, firmware development, mobile app creation, and statistical validation.

---

License

MIT License — see LICENSE

---

Citation

If you use this project in research or teaching:

@software{bkshoot2021,
  author = {Moros Vilaseca, Oriol},
  title = {BK-Shoot: Arduino + MIT App Inventor Basketball Shot Tracker},
  year = {2021},
  url = {https://github.com/omorros/bk-shoot}
}

See CITATION.cff for full citation information.