Headcount

Pitch_PDF_Headcount.pdf has been include as a concise pitch deck (not a presentation). Application recording has been included as a Youtube link.

Inspiration

Chirag Patel, a hackathon mentor, and several other participants got stuck in elevators recently.

What it does

1. Project Title:

   • Headcount: Ensuring capacity limit compliance for elevator safety and efficiency by using real-time vision AI to prevent overcrowding, and enhance the publics experience.

2. Team Members:

   Frank Buibish - Developer

   DJ Nadgar - Product Management

   David Zarco - Design

3. Problem Set:

- General AI Track Problem 1: Smart CIty

1. Project Description:

- A brief overview of your project:

    Elevators in high-traffic buildings often face issues of overcrowding, which can lead to several problems including: safety hazards, operational inefficiency and bad user experience. 

    How might we ensure the safety and comfort of elevator passengers by effectively monitoring and managing elevator capacity in real-time?

    **Solution:** Headcount is an innovative elevator safety app designed to address the problem of overcrowding by utilizing real-time video analytics and AI. Real-time person head-counting is utilized which notifies the user or emergency responders so they may take action. 

    **Potential Impact:** Market Size: The global CCTV camera market is projected to grow from $35.47 billion in 2022 to $105.20 billion by 2029, at a CAGR of 16.8% in forecast period, 2022-2029

    In addition Headcount can move into other avenues such at public infrastructure (buses, Metro mover, Metrorail) and other large crowd areas like piers, clubs or festivals.  

1. Technologies Used:

- React/Next.js
- Linux/Python
- Figma
- Google Vision/FFMpeg
- Twilio
- Google Cloud CLI

1. Current Progress:
   • Describe the current state of your project, including:
     • Key features implemented so far.
     • We can detect faces/heads in real time from a Youtube stream and we can generate Twilio alerts via WhatsApp with the face/head count from that detection or change in detection.
     • Separate Figma prototype of more realistic UI
     • Any challenges or obstacles you have encountered.
     • Inability to livestream to test the product because we need 50 subscribers and for 24 hours advance on Youtube for the first livestream. This hurdle was overcome by using a livestream from someone with more than 440 subscribers who attended the Hackathon. DONE
     • Next steps you plan to take.
     • Record a demo to show the code working; detection and notification. DONE
     • Target of opportunity, create the UIs detailing in the Figma prototype. DONE
2. Preliminary Screenshots or Demos:
   • Include any initial screenshots, sketches, or video demos that showcase your progress.

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for Headcount

Built With

• figma
• gcp
• google/vissionai
• linux/python
• next.js
• react
• twilio
• vision/ffmpeg