How to Connect TFT LCD (ILI9341, ST7789) to AI Agent ASI Biont: IoT Dashboard Setup Guide

Why Integrate a TFT LCD with an AI Agent?

TFT LCD displays based on ILI9341 (320×240) and ST7789 (240×240, 135×240) are ubiquitous in IoT projects—from weather stations and smart home dashboards to industrial panel meters. Typically, you write custom firmware in C++ (Arduino) or MicroPython to update a screen with sensor data. But what if your display could be controlled by an AI agent that makes decisions, sends alerts, and even generates new UI layouts on the fly?

With ASI Biont, you don't need to manually code every pixel update. The AI agent connects to your microcontroller (ESP32, Raspberry Pi Pico, STM32) via MQTT or Hardware Bridge (COM port) and sends commands to display real-time IoT metrics, AI-generated notifications, and dynamic dashboards—all through a natural language chat.

Connection Methods: MQTT vs. Hardware Bridge

For this article, we'll focus on the MQTT approach—most flexible for remote dashboards. The AI agent (running in ASI Biont's cloud) uses paho-mqtt to publish display commands to an ESP32 that drives an ILI9341 TFT.

Real-World Use Case: AI-Powered IoT Dashboard

Scenario: You have an ESP32 with a DHT22 temperature/humidity sensor and a 2.8" ILI9341 TFT. You want the display to show:
- Current temperature, humidity, and time
- Color-coded alerts (red if temp > 30°C)
- A simple bar chart of the last 24 hours of readings

Traditionally, you'd write hundreds of lines of Arduino code to parse MQTT messages, draw fonts, and update graphics. With ASI Biont, you describe the task in chat:

"Connect to my ESP32 via MQTT at 192.168.1.100:1883, topic 'display/cmd'. The ESP32 runs a MicroPython script that subscribes to that topic and renders text and charts on an ILI9341 TFT. I want to show current temp, humidity, and a red alert when temp exceeds 30°C. Also, every 10 minutes, publish a command to draw a simple bar chart of the last 6 readings."

The AI agent then:
1. Generates the MicroPython firmware for the ESP32 (using st7789 or ili9341 library)
2. Generates a Python script that runs in ASI Biont's sandbox (via execute_python), subscribes to sensor data, and publishes display commands via MQTT
3. Starts the integration automatically

Step-by-Step: Code Examples

Step 1: ESP32 Firmware (MicroPython)

The ESP32 runs this minimal subscriber. It listens on display/cmd and updates the TFT:

from machine import Pin, SPI
import st7789
import ujson
import time
import network
from umqtt.simple import MQTTClient

# Initialize display
tft = st7789.ST7789(
    SPI(1, baudrate=40000000, sck=Pin(18), mosi=Pin(23)),
    240, 240,
    reset=Pin(33, Pin.OUT),
    cs=Pin(5, Pin.OUT),
    dc=Pin(19, Pin.OUT),
    backlight=Pin(4, Pin.OUT),
    rotation=1
)
tft.init()
tft.fill(st7789.BLACK)

# MQTT callback
def mqtt_callback(topic, msg):
    data = ujson.loads(msg)
    if "text" in data:
        tft.text(data["font"], data["text"], data["x"], data["y"], data["color"])
    elif "fill" in data:
        tft.fill(data["color"])
    elif "rect" in data:
        tft.rect(data["x"], data["y"], data["w"], data["h"], data["color"])
    elif "hline" in data:
        tft.hline(data["x"], data["y"], data["w"], data["color"])

# Connect to Wi-Fi
wlan = network.WLAN(network.STA_IF)
wlan.active(True)
wlan.connect("SSID", "PASSWORD")
while not wlan.isconnected():
    time.sleep(1)

# Connect to MQTT
client = MQTTClient("esp32", "192.168.1.100")
client.set_callback(mqtt_callback)
client.subscribe(b"display/cmd")
client.connect()

while True:
    client.wait_msg()

Step 2: AI Agent Python Script (Runs in ASI Biont Sandbox)

This script runs in execute_python on the ASI Biont server. It reads sensor data (simulated here), analyzes it, and publishes display commands via MQTT:

import paho.mqtt.client as mqtt
import json
import random
import time
from collections import deque

broker = "192.168.1.100"
topic = "display/cmd"
client = mqtt.Client()
client.connect(broker)

history = deque(maxlen=6)

def update_display(temp, humidity):
    # Clear screen
    client.publish(topic, json.dumps({"fill": 0x0000}))

    # Draw header
    client.publish(topic, json.dumps({"text": "ASI Biont Dashboard", "x": 10, "y": 10, "color": 0xFFFF, "font": "small"}))

    # Temperature with color
    color = 0xF800 if temp > 30 else 0x07E0
    client.publish(topic, json.dumps({"text": f"Temp: {temp:.1f}C", "x": 10, "y": 40, "color": color, "font": "large"}))

    # Humidity
    client.publish(topic, json.dumps({"text": f"Humidity: {humidity:.1f}%", "x": 10, "y": 70, "color": 0x07FF, "font": "medium"}))

    # Bar chart (if enough data)
    if len(history) == 6:
        max_val = max(history)
        bar_width = 30
        for i, val in enumerate(history):
            bar_height = int((val / max_val) * 100)
            x = 10 + i * (bar_width + 5)
            y = 180 - bar_height
            client.publish(topic, json.dumps({"rect": x, "y": y, "w": bar_width, "h": bar_height, "color": 0x001F}))

# Simulate sensor readings
for _ in range(10):
    temp = random.uniform(20, 35)
    hum = random.uniform(40, 70)
    history.append(temp)
    update_display(temp, hum)
    time.sleep(5)

How the User Connects Everything

1. Flash the ESP32 with the MicroPython firmware (using esptool or Thonny).
2. Run bridge.py on a PC (optional) if using COM port instead of MQTT.
3. Open a chat with ASI Biont and describe your setup:

   "I have an ESP32 with an ST7789 TFT at IP 192.168.1.100, MQTT broker at port 1883. Topic for display commands is 'display/cmd'. Publish temperature and humidity every 10 seconds, with red alerts above 30°C."

4. The AI agent writes and executes the integration code instantly.

Why This Approach Wins

• Zero manual coding: The AI generates both the microcontroller firmware and the server-side logic. You only describe the behavior.
• Dynamic dashboards: Change the layout, add new metrics, or switch fonts by simply telling the AI—no re-flashing required.
• Real-time AI decisions: The AI can analyze sensor trends and push alerts (e.g., "temperature spike detected—display warning") without human intervention.

Conclusion

Integrating a TFT LCD (ILI9341/ST7789) with ASI Biont turns a simple display into an AI-driven control panel. Whether you're building a smart thermostat, a factory floor monitor, or a weather station, the AI agent handles all the protocol plumbing—MQTT, COM port, or SSH—and generates production-ready code in seconds.

Ready to give your TFT LCD a brain? Try the integration at asibiont.com and see how AI can automate your next IoT dashboard.