Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 

Repository files navigation

NBN Sorter - Automated Bolt, Nut & Washer Sorting Machine

Design

A mechatronics project by students of **Kasetsart University Sriracha Campus**
Industrial Engineering Program

NBN

Project Overview

NBN Sorter is an automated mechanical sorting machine designed to separate mixed hardware fasteners — bolts, nuts, and washers — by type and size (M3, M4, M5) without human intervention.

The system uses a combination of gravity-based rail sorting, magnetic conveyance, rotating disk mechanisms, and a real-time TFT dashboard to classify and count each fastener type.

Input: Mixed bolts (M3/M4/M5), nuts, and washers poured into the input hopper
Output: Sorted fasteners separated into individual collection bins by size and type

Overview

System Architecture

The machine consists of 4 stations connected in sequence, plus a display/wiring unit:

Workflow

Station Details

Station 1 - Rail Sorter

Station 1-1

Station 1-2

Station 1-3

  • Mixed fasteners are poured at the top-right input hopper

  • Gravity guides bolts along one path to Station 2

  • Nuts and washers slide down to separate sorter rails

  • An adjustable gap mechanism separates by thickness:

    • 1.2 mm gap → washers pass through
    • 3.8–4.1 mm gap → nuts pass through
  • An acrylic cover guide prevents parts from bouncing off due to motor vibration

  • Spring-loaded coupling increases incline force to prevent jamming

Accuracy: 85% (255/300 over 10 test runs)

Station 2 - Pink Box (Bolt Feeder)

Station 2-1

  • Receives bolts from Station 1
  • Internal rotating disk with hidden magnetic buttons picks up bolts individually
  • Bolts are lifted and guided along a rail to Station 4
  • Magnetic hole offset: 2.0 mm (optimal for single-bolt pick-up)
  • Guide rail incline was increased to reduce friction-related jamming

Accuracy: 39% (39/100 over 10 test runs) — lowest performing station; main bottleneck

Station 3 - Rotating Plate (Nut Sorter)

Station 3-1

Station 3-2

Station 3-3

  • Receives nuts from Station 1

  • A rotating disk with a fixed semi-circle plate directs nuts into size-based curve paths

  • Sorting gap parameters (distance between rotating plate and fixed semi-circle):

    • M3 section: 4.4 mm
    • M4 section: 3.5 mm
    • M5 section: 2.8 mm
  • Drive system: bevel gear reduction at 1:4.5 ratio (replaced original belt-pulley drive)

  • Front support added to the semi-circle plate for perpendicular stability

Accuracy: 97.6% (117/120 over 10 test runs) — best performing station

Station 4 - Magnetic Conveyor (Bolt Sorter)

Station 4-1

  • Receives bolts from Station 2
  • Bolts attach to magnets embedded in the conveyor rail surface
  • As the belt moves, bolts drop into M3/M4/M5 collection boxes below
  • Wall thickness between bolt and magnet: 1.5 mm offset (optimal release force)
  • Gear ratio: 1:3
  • Conveyor belt uses custom-designed PLA linkage (replaced thin metal wire)
  • Rail designed from scratch in SolidWorks with embedded magnet holes

Accuracy: 64% (64/100 over 10 test runs)

Display & Wiring — NBN Dashboard

  • Built on ESP32 with TFT eSPI display
  • GUI rendered using LVGL library
  • Displays real-time count of sorted bolts (M3/M4/M5) and nuts/washers
  • Serial communication with sensors via circular buffer
  • Dynamic label updates via updateUI() function based on sensorOut\[] array

Overall Results

Station Component Accuracy
Station 1 Rail Sorter 85%
Station 2 Pink Box 39%
Station 3 Rotating Plate 97.6%
Station 4 Magnetic Conveyor 64%
Overall System End-to-end (3 runs) 40.95%

End-to-End Test Summary (3 runs)

Size Bolt Nut Washer
M3 26.6% 40%
M4 3.3% 40%
M5 10% 80%
OTHER 90%
Category Total 35.71% 30% 57.14%

Overall system accuracy: **40.95%**

Known Issues

  • Station 2 (Pink Box) is the primary bottleneck — magnetic clumping causes multiple bolts to jam together
  • Washer sensor reads 0% — washers could not be detected reliably in the dashboard
  • Bolt detection in the dashboard is slow and error-prone
  • PLA deformation requires manual filing with vernier caliper to maintain tolerances
  • End-to-end accuracy is limited by inter-station transfer reliability

Design Iterations

The machine went through 6 design versions before reaching the final build, iterating on:

  • Structural layout and station arrangement
  • Drive systems (belt-pulley → bevel gear reduction)
  • Conveyor belt material (metal wire → PLA linkage)
  • Sorting gap dimensions (empirically tested)

Team Member

  • Chanaprachpakorn Ngernmo
  • Theerapol Guanmuangtai
  • Thanakrit Wongkha
  • Warut Tongrung
  • Norawich Lorttrakanont
  • Kunasin Laysak
  • Danuphat Parnpradit
  • Thanakon Chaichana
  • Jerich Brylle Sison
  • Sajakorn Chutikarnmongkol

Institution

Kasetsart University Sriracha Campus
Industrial Engineering Program

This project was developed as part of an academic engineering design course. Full Demonstration Video: 🔗 Watch on YouTube

About

The automated mechanical sorting machine designed to separate bolts, nuts, and washers by type and size without human intervention.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages