Projects

Objective

• • There is no way to know electronically when certain rooms/floors of public space (such as a college library or lounge) are occupied, which makes it hard to plan the use of a certain space ahead of time.

Background

• • Create an electronic sensor system that monitors the real-time occupancy of a building, which information can be uploaded and visualized on a floor plan for students to more efficiently plan where to go.

Methodology

• • Sensors can be set up spaced around an entire building, which will be able to detect motion or noise in a given time interval. This data will be published to a server for use in the application, which will visualize the "hot-spots" of a given floor within the building. Given where these hot-spots are, the application will be able to prompt users where the least occupied areas are in the building.

Expected Results

• • Students will be able to track the capacity and congestion of certain spaces electronically, either from their phone or laptop.

Costs

• Raspberry Pi (Have several already)

• PIR Motion Sensor (~$14)

• Arduino Voice Detection Module (~$30)

• Software (Free)

• Other Electrical Components - Wires, etc... (~$10)

• Stated Problem: Improve ways to be able to track the congestion and occupancy of a space
• Real Problem: Install measures to track activity such as noise and motion in a space, which is more effective in conveying how occupied a certain space is at a moments notice

• Given/Inferred boundaries: The cost of sensors and equipment being high, as well as the setup of IoT in real-time being difficult to implement.
• Actual Boundaries: Privacy and complaints with using motion and noise sensing equipment.

• Given/Inferred Goals: Use IoT to map congestion in public spaces.
• Meaningful Goals: Connect various external sensing mechanisms including a microphone and PIR sensor to a Raspberry Pi, and use real-time data to create a 'map' of hotspots.

• What kind of sensory input will be the most useful to help analyze how spaces are occupied at given times -> Relate this to find the best possible output.
• Unknowns - One such problem that has appeared is that of privacy in the given spaces, and how to work around this issue while still getting the best possible outcome in terms of data.

• Start by placing Raspberry Pis in classrooms/public spaces on college campuses
• Have tracking capabilities for each Raspberry Pi to record time people enter a room by detecting motion changes
• Connect to central web-server
  • Have simple GUI for viewing purposes for college students to find open spaces throughout the day
• Be discrete with little hinderance to users using the product

• User friendly Graphical User Interface
• Minimize delay and possibilities of errors being recorded
• Easy to install/set up in rooms

• Detect when users enter or leave
• High power efficiency and ease of installation to minimize costs/error

• Determine if one or multiple people pass through the motion sensor.
• Reliable detection of people

• Raspberry Pi
• Motion Sensors
• Web-server
• Mobile-application written in Java for Android and Swift for iPhone
• OpenCV code to detect people in images

• Both sound and motion would help achieve the detection of people but sound would be too unreliable. Someone might be in a lounge quietly studying and other people could be in there having a conversation. It would be difficult to pinpoint the quiet person in the loud group.