Here’s your refined and clarified Devpost version — polished for clarity, flow, and impact while keeping the authentic hackathon tone:


🧠 ASL Express – Touchless AI Food Ordering Assistant

💡 Inspiration

For many deaf and mute individuals, a simple trip through a drive-thru can be frustrating, stressful, or even impossible. In fast-paced and noisy environments, communication barriers often turn everyday tasks—like ordering food—into overwhelming experiences.

We wanted to change that. ASL Express gives users a voice through vision. Using AI and computer vision, our system translates hand signs into accurate, real-time food orders. It’s not just about convenience—it’s about inclusion, independence, and dignity.

But our vision extends beyond drive-thrus. ASL Express can empower people in restaurants, hospitals, schools, airports, and public kiosks, creating a truly universal, touchless ordering experience. What started as a hackathon project could easily grow into a scalable startup solution, helping millions communicate effortlessly every day.


⚙️ What it does

ASL Express is a touchless AI-powered ordering system that uses a laptop camera to recognize hand gestures and convert them into food orders.

Each sign corresponds to a menu item:

  • A → Burger 🍔
  • B → Fries 🍟
  • C → Drink 🥤

The number of repetitions (1–3) represents quantity, and a thumbs-up gesture finalizes the order.

Once confirmed, the system sends the order to an ESP32 microcontroller, which displays it on an LCD screen, activates LEDs for visual feedback, and triggers a buzzer for confirmation—creating an intuitive, multi-sensory interaction loop.


🛠️ How we built it

  • Developed the gesture recognition pipeline using Python, MediaPipe, and Google Gemini API for classification and intent mapping.
  • Designed and programmed ESP32 hardware integration including LCD, LEDs, ultrasonic sensor, and buzzer.
  • Established serial communication between Python and ESP32 to send recognized commands in real time.
  • Built a feedback system using light and sound cues to confirm recognition and ensure accessibility for all users.

🚧 Challenges we ran into

  • Synchronizing real-time gesture recognition with hardware response.
  • Calibrating Gemini’s AI outputs with MediaPipe hand landmarks for consistent accuracy.
  • Handling serial latency and preventing false gesture detections.
  • Managing power and pin limitations on the ESP32 while connecting multiple peripherals.

🏆 Accomplishments we’re proud of

  • Built a fully functional prototype that bridges AI vision and embedded hardware.
  • Achieved reliable recognition for four gestures plus a “Done” signal with high accuracy.
  • Designed a beginner-friendly, inclusive interface that works in real time.
  • Combined software, AI, and hardware engineering seamlessly under hackathon constraints.

📚 What we learned

  • Integrating AI APIs (Gemini) into embedded IoT systems.
  • Deep understanding of gesture tracking and hand landmark detection.
  • Building serial communication protocols for synchronized multi-device interaction.
  • Enhancing teamwork, adaptability, and rapid prototyping within tight deadlines.

🚀 What’s next for ASL Express

  • Expand gesture support to full ASL alphabet recognition for a richer vocabulary.
  • Add voice feedback and confirmations using ElevenLabs API.
  • Deploy touchless AI kiosks in restaurants, hospitals, and schools.
  • Explore AI camera modules for on-device gesture recognition, reducing dependency on PCs.

Built With

  • api
  • buzzer-platforms-&-tools:-arduino-ide
  • c++-(for-esp32-firmware)-frameworks-&-libraries:-mediapipe
  • code
  • communication
  • elevenlabs
  • languages:-python
  • lcd-display
  • leds
  • opencv
  • other
  • protocols:
  • pyserial-apis-&-ai-models:-google-gemini-api-(for-gesture-analysis-and-reasoning)-hardware:-esp32
  • serial
  • services:
  • studio
  • ultrasonic-sensor
  • usb/uart)
  • visual
Share this project:

Updates