Adaptive Traffic Signal Control System for Real-Time Optimization of Vehicular Flow at Urban Intersections

Project Overview

Traditional fixed-cycle traffic lights waste time on empty roads and create unnecessary congestion on busy approaches. This system replaces that rigid pattern with real-time density-based control: two HC-SR04 ultrasonic sensors per lane (front and back) measure vehicle queue length continuously, and each lane is classified as EMPTY, NORMAL, or CROWDED. Green-light duration is then allocated dynamically — 0 s (Smart Interrupt), 15 s (normal), or 30 s (crowded) — so no lane idles unnecessarily while vehicles wait on the opposing axis.

The firmware runs on an Arduino Mega 2560 using a non-blocking millis()-based 6-state machine. This architecture allows the microcontroller to simultaneously scan all eight sensors every 300 ms, update the 20×4 I2C LCD countdown display every 250 ms, manage yellow-phase transitions, and respond instantly to the hardware emergency override button — all without a single delay() call in the main loop.

Key Features

• Dual-sensor lane detection: front + back HC-SR04 per lane (8 sensors total); both must clear 12 cm for CROWDED, either one for NORMAL, neither for EMPTY
• Worst-case axis prioritisation: each axis takes the maximum density of its two constituent lanes before allocating green time
• Smart Interrupt: collapses active green phase instantly when current axis is EMPTY and opposing axis detects vehicles
• Hardware Emergency Override: debounced push button on Pin 12 (INPUT_PULLUP); toggles all lanes to RED and flashes consoleLED at 200 ms; LCD shows emergency message; press again to resume
• Non-blocking millis() architecture: sensors scanned every 300 ms, LCD refreshed every 250 ms, emergency debounce at 50 ms — zero delay() calls
• 20×4 I2C LCD dashboard: live active axis, traffic density classification, phase mode (GREEN/YELLOW), and countdown timer in seconds
• Safety yellow transition: mandatory 5 s yellow phase before every green-to-red switch across both axes
• Splash screen on startup: displays developer name, matric number, and project affiliation for 5 s before entering autonomous control

System Architecture

The system is organised as a single non-blocking loop() that services five concurrent tasks:

Hardware Components

Arduino Mega Pin Assignments

Ultrasonic Sensor Pins

LED and Control Pins

Circuit Diagrams

Both diagrams were produced in Proteus Design Suite and show the full 4-lane intersection wiring: Arduino Mega 2560, 8 HC-SR04 sensors in 4 front-back pairs, 12 traffic LEDs with 220 Ω current-limiting resistors, 20×4 I2C LCD, emergency push button, and the 12 V PSU with buck converter regulated 5 V rail.

Colour Proteus schematic: full system wiring with teal sensor harnesses, LED resistor networks, I2C LCD, and PSU/buck converter rail

Monochrome Proteus schematic: identical layout in black and white for print and documentation clarity

Traffic Signal State Machine

The controller cycles through six states in a fixed sequence. Axis 1 (Lanes 1 & 4, top and bottom) and Axis 2 (Lanes 2 & 3, left and right) alternate green phases separated by mandatory 5 s yellow transitions. Both GREEN states include a Smart Interrupt check on every loop tick.

Firmware

Written in Arduino C++ (Arduino IDE) using the HCSR04 library for sensor abstraction and LiquidCrystal_I2C for LCD control. Three key excerpts are shown below.

Density Classification

// Returns EMPTY, NORMAL, or CROWDED based on front and back sensor readings.
// Negative readings (no echo) are clamped to 999 cm = EMPTY.
Density getLaneDensity(float frontCm, float backCm) {
  if (frontCm < 0) frontCm = 999;
  if (backCm  < 0) backCm  = 999;

  if (frontCm < 12.0 && backCm < 12.0) return CROWDED;
  if (frontCm < 12.0 || backCm < 12.0) return NORMAL;
  return EMPTY;
}

void scanAllLanes() {
  Density L1 = getLaneDensity(L1_Sens1.measureDistanceCm(), L1_Sens2.measureDistanceCm());
  Density L4 = getLaneDensity(L4_Sens1.measureDistanceCm(), L4_Sens2.measureDistanceCm());
  axis1Status = max(L1, L4);  // worst-case axis density

  Density L2 = getLaneDensity(L2_Sens1.measureDistanceCm(), L2_Sens2.measureDistanceCm());
  Density L3 = getLaneDensity(L3_Sens1.measureDistanceCm(), L3_Sens2.measureDistanceCm());
  axis2Status = max(L2, L3);
}

State Machine Core with Smart Interrupt

void manageTrafficLogic() {
  unsigned long elapsed = millis() - stateStartTime;

  switch (currentState) {

    case AXIS1_EVAL:
      activeAxisStr = "1 & 4 (MAIN)";
      currentDuration = (axis1Status == CROWDED) ? TIME_CROWDED
                      : (axis1Status == NORMAL)  ? TIME_NORMAL
                      :                            TIME_EMPTY_OSCILLATE;
      triggerAxis1Green();
      currentState = AXIS1_GREEN;
      break;

    case AXIS1_GREEN:
      // Smart Interrupt: skip idle green if opposing axis has vehicles
      if (axis1Status == EMPTY && axis2Status > EMPTY) { currentDuration = 0; }
      if (elapsed >= currentDuration) { triggerAxis1Yellow(); currentState = AXIS1_YELLOW; }
      break;

    case AXIS1_YELLOW:
      if (elapsed >= TIME_YELLOW) { currentState = AXIS2_EVAL; }
      break;

    case AXIS2_EVAL:
      activeAxisStr = "2 & 3 (SIDE)";
      currentDuration = (axis2Status == CROWDED) ? TIME_CROWDED
                      : (axis2Status == NORMAL)  ? TIME_NORMAL
                      :                            TIME_EMPTY_OSCILLATE;
      triggerAxis2Green();
      currentState = AXIS2_GREEN;
      break;

    case AXIS2_GREEN:
      if (axis2Status == EMPTY && axis1Status > EMPTY) { currentDuration = 0; }
      if (elapsed >= currentDuration) { triggerAxis2Yellow(); currentState = AXIS2_YELLOW; }
      break;

    case AXIS2_YELLOW:
      if (elapsed >= TIME_YELLOW) { currentState = AXIS1_EVAL; }
      break;
  }
}

Emergency Override

void checkEmergencyButton() {
  int reading = digitalRead(buttonPin);
  if (reading != lastButtonState) { lastDebounceTime = millis(); }
  if ((millis() - lastDebounceTime) > 50) {
    if (reading != buttonState) {
      buttonState = reading;
      if (buttonState == LOW) {
        emergencyMode = !emergencyMode;
        if (emergencyMode) { setAllRed(); lcd.clear(); }
        else               { lcd.clear(); }
      }
    }
  }
  lastButtonState = reading;
}

void handleEmergencyMode() {
  // Flash consoleLED at 200 ms
  if (millis() - lastFlashTime >= 200) {
    lastFlashTime = millis();
    flashState = !flashState;
    digitalWrite(consoleLED, flashState ? HIGH : LOW);
  }
  // Update LCD every 1 s
  if (millis() - lastLcdUpdate >= 1000) {
    lastLcdUpdate = millis();
    lcd.setCursor(0, 0); lcd.print("!!! EMERGENCY !!!   ");
    lcd.setCursor(0, 1); lcd.print("ALL LANES HALTED    ");
    lcd.setCursor(0, 2); lcd.print("WAITING TO RESUME...");
    lcd.setCursor(0, 3); lcd.print("PRESS BTN TO CANCEL ");
  }
}

Build Gallery

134 photographs documenting the complete build process from initial component procurement through to final system presentation.

Components and Procurement — January 2026

Cleared workbench surface ready for component layout

DC-DC step-down (buck) converter module — angle 1

DC-DC step-down (buck) converter module — angle 2

DC-DC step-down (buck) converter module — angle 3

HC-SR04 ultrasonic distance sensor modules — batch 1

HC-SR04 ultrasonic distance sensor modules — batch 2

9×15 cm copper perfboard for PSU circuit assembly

20×4 character LCD display module

Arduino Mega 2560 in original retail packaging

Arduino Mega 2560 board — close-up showing digital pin banks used for sensor and LED wiring

Solder wire reel for permanent perfboard connections

Rocker power switch for control box — angle 1

Rocker power switch for control box — angle 2

I2C LCD backpack module (PCF8574 at address 0x27) — simplifies LCD to two-wire I2C

Red, yellow, and green 5 mm LEDs — 4 sets for the traffic signal outputs across all lanes

Female pin header strips for detachable sensor and LED connections

Tactile push button for emergency override — side view

Tactile push button for emergency override — top view (Pin 12, INPUT_PULLUP)

220 Ω resistors for traffic LED current limiting (12 required)

BC547 transistors and passive buzzer components

Transistors and buzzer components — overhead view

Male-to-female Dupont jumper wires for prototyping connections

Ribbon cable used for multi-sensor wiring harnesses from sensor bars to Arduino

Single-core hook-up wire for permanent soldered connections on perfboard

12 V DC mains power adapter — primary supply for the buck converter rail

Power supply unit interior showing mains-to-DC conversion wiring

Power Supply Setup and Sensor Preparation — February 2026

Buck converter wired into the 12 V power supply rail for regulated 5 V output

Multimeter verifying regulated 5 V output from buck converter — reading 1

Voltage output verification with multimeter — reading 2

Voltage output verification — reading 3

Voltage output verification — reading 4, confirming stable 5 V before wiring to Arduino

Hand-drawn circuit schematic sketched in notebook before moving to Proteus for final design

PSU and buck converter module mounted and soldered permanently on perfboard

HC-SR04 sensors wired together via ribbon cable for sensor bar assembly — angle 1

HC-SR04 sensors wired via ribbon cable — angle 2

20×4 I2C LCD powered and connected — confirming I2C address 0x27 with the LiquidCrystal_I2C library

LED signal cable harness assembly — angle 1 (12 LED wires pre-cut and stripped)

LED signal cable harness assembly — angle 2

Enclosure Material Preparation — February 2026

Composite board panels procured for the model base and control box enclosure — outdoor staging

Composite board elevated for sizing assessment before cutting

Composite boards after initial cutting to required panel dimensions

Cut composite panels leaning against wall after sizing

Composite board panel with nail positions marked for enclosure box assembly

Board layout and measurement planning for the intersection model base — view 1

Board layout and measurement planning — view 2

Composite board with box frame attached for control box enclosure structure

Electronics workbench during early enclosure preparation — angle 1

Electronics workbench during early enclosure preparation — angle 2

Drilling workbench set up with tools for cutting panel openings and component cutouts

Control panel board with cutouts routed for rocker switch, emergency button, and cable ports

Early circuit boards test-mounted on the base board for component layout and spacing validation

Base Board and Sensor Bar Construction — March 2026

Black acrylic sheet and PVC pipes sourced for the road surface panels and base elevation legs

MDF base board with PVC stub legs fitted — elevates the model for underside wiring access

Wood-plastic composite (WPC) perimeter frame construction — angle 1

WPC perimeter frame construction — angle 2

Overhead view of completed WPC perimeter frame before road surface and green panel installation

MDF base board with the 4-lane cross-intersection road grid pencilled in for cutting reference

Detailed road grid with lane widths and intersection boundaries measured and marked on MDF base

Composite board stack with metal corner brackets ready for enclosure box assembly

Cardboard template laid out to plan road intersection shape and sensor placement — angle 1

Cardboard template for road intersection layout — angle 2

PVC cable trunking strips arranged loosely to plan road lane wiring channel routing

Trunking intersection plan showing the 4-lane cross layout for concealed wiring beneath the road surface

Stack of cable trunking strips cut to road-lane dimensions

Trunking strips with sensor mounting holes and wire exit slots cut out

Sensor bar with four HC-SR04 sensors mounted (two lanes, front and back each) — angle 1

Sensor bar with four HC-SR04 sensors — angle 2

Sensor bar with four HC-SR04 sensors — angle 3

Sensor bar with four HC-SR04 sensors — angle 4

Completed dual-sensor bar with ribbon cable harness ready for lane installation

Road Surface and Model Assembly — March 2026

MDF intersection board fitted and secured within the WPC perimeter frame

Sensor bars dry-run positioned across lane entry points to check spacing and alignment

Green acrylic sheet being cut to form the road verge (grass) panels flanking each lane

Sensor bars positioned within green acrylic verge surround — angle 1

Sensor bars positioned within green acrylic verge surround — angle 2

Green acrylic verge panels partially installed revealing the 4-lane road cross layout — angle 1

Green acrylic verge panels partially installed — angle 2

Green verge panels and sensor bars fully fitted — full 4-lane cross intersection visible

Black acrylic road surface taking shape at the intersection centre

Intersection road surface further developed with lane approach strips laid in

Ruler confirming intersection road dimensions and lane widths for sensor alignment accuracy

Drilling LED mounting holes along each lane stop-line position

Step drill bit boring LED mounting holes in the road surface panel — view 1

Step drill bit boring LED holes — view 2

Step drill bit boring LED holes — view 3 showing completed hole size

Painted pedestrian crosswalk markings applied to the road surface — view 1

Painted pedestrian crosswalk markings — view 2 showing all four crossings at the intersection

Completed road surface: crosswalk markings painted, LED holes drilled, ready for electronics integration

Developer with Assembled Road Model — March 2026

Developer with assembled road model — selfie 3

Developer portrait with road model — portrait 2

Developer with road model showing traffic light positions — view 2

Developer with road model showing traffic light positions — view 3

Developer with road model showing traffic light positions — view 4

Electronics Integration and Full System Wiring — March 2026

Road model alongside the Arduino Mega control box before final wiring integration

Overhead view of model with control box and wiring harnesses routed — angle 1

Overhead view of model with control box — angle 2

Full overhead view of the completed intersection model: 4 lanes, LED traffic lights at each stop-line, sensor bars at entry points, and crosswalk markings

Close-up of traffic light LEDs installed in road model stop-line holes — red, yellow, and green per lane

Developer Jemimah Asma’u Kyauta Thompson with the fully wired traffic signal system — photo 1

Developer with fully wired system — photo 2

Developer with fully wired system — photo 4

Developer with fully wired system — photo 6

Developer with fully wired system — photo 7

Developer with fully wired system — photo 8

Developer with fully wired system — photo 9

Developer with fully wired system — photo 10

Developer with fully wired system — photo 11

Developer with fully wired system — photo 12

Developer with fully wired system — photo 13

System integration and testing session — full model and control box laid out on the floor for wiring verification

Final road model — top view showing completed intersection with all LEDs and sensor bars installed

Underside of model showing PVC stub legs and sensor/LED wiring routed neatly to the control box

Arduino Mega control box interior — angle 1 showing Mega board, perfboard PSU, and incoming wiring harnesses

Arduino Mega control box interior — angle 2 showing full wiring layout

Control box with 20×4 I2C LCD mounted and powered — LCD displaying live system status — angle 1

Control box with 20×4 I2C LCD — angle 2 showing active axis, density, and countdown timer

Completed System and Final Presentation — May 2026

Developer full-length portrait at final presentation

Developer with the completed adaptive traffic signal system — view 1

Developer with the completed adaptive traffic signal system — view 2

Completed system in night-time operation — illuminated red, yellow, and green LEDs visible across all four lanes

Developer holding the completed model for presentation — showing full model scale

Developer presenting model from overhead — full 4-lane cross intersection visible with green verge panels, sensor bars, and LED stop-lines

Project Links

Thank You for Visiting My Portfolio

I sincerely appreciate you taking the time to explore my portfolio and learn about my work and expertise. If you have any questions or wish to discuss potential collaborations, please feel free to reach out via the Contact section.

Best regards,
Damilare Lekan, Adekeye.