BTDS-project

Room Monitoring Authors: Maximilian Busch Felix Niebl

Goal

Implementation of a System that can create an internal model of a building complex at room scale resolution and track the amount of people in each room within reasonable accuracy. Additionally, the System to be developed needs to be able to track the distance of inhabitants to each other at certain areas of interests and ring an alarm in case a minimum safety distance is not met.

To reach this Goal we created a System containing 3 components:

1. Room surveillance sensor (RSS)
   A sensor system that consistently observes an area of interest in a given room and sends the positions of it's inhabitants to the central server.

2. Passageway Surveillance sensor (PSS)
   A sensor system that consistently observes a passageway or an entrance to a room and notifies the server whenever a person enters or leaves a room.

3. Central Tracking Server (CTS)
   The Central server, which is tasked with keeping track of the inhabitant numbers of each room by using information from an arbitrary amount of both types of sensor systems.

installation requirements

Each instance of an RSS and PSS will require a kinect V2 sensor and a separate windows machine that is capable of supporting such Sensors. The CTS can run on any system that is capable of supporting the used components: nodeJS and mongoDB. Generally it is possible to run the CTS on one of the sensor-machines. All machines used need to be able to communicate with the CTS-host-machine via a network connection.

installation instructions

CTS installation

Installation

1. Install the nodeJS runtime environment on your computer: https://nodejs.org/en/
2. Install the mongoDB community server on the same computer: https://www.mongodb.com/try/download/community (it may also be possible to use a server on a different machine or in a cloud. It will however need a different address)
3. Once ensured that these systems are running, clone the project repository https://github.com/niebl/BTDS-project
4. Use a CLI of your choice to navigate to the CTS directory and run “> npm install” (if this doesn’t work, check if node is added to the path variables)
5. Make a copy of the file “.env_sample” and rename it to “.env”. Open it in an editor and change the variables if you need to.
6. Start the server using “npm start”.
7. Open your browser and visit “localhost:3000”, to see if the server is working (or any other port if you have changed the .env).
8. You may also view the page on a mobile device within the same network by visiting “{server-IP-address}:3000” in the browser of the device. This can be useful when testing the system.

If the page doesn’t load and the console returns “UnhandledPromiseRejectionWarning: MongooseError: Operation”, make sure the mongoDB server is running and accessible over the address that was specified in the .env.

installation PSS and RSS

(also referred to as Sensor-System)

1. On each sensor-system, configure and install the Kinect V2 depth-Sensor. For this, refer to this guide: https://github.com/violetasdev/bodytrackingdepth_course/wiki/Kinect-V2.
2. To have a tool to build the sensor-system-applications from the repository, install and configure Visual Studio 2019 for the kinect V2. For this, refer to this guide: https://github.com/violetasdev/bodytrackingdepth_course/blob/master/KinectV2/docs/visualStudio2019_doc.md.
3. Clone the project repository and navigate to the directory “PassagewaySensorSystem” or “RoomSensorSystem” respectively, depending on which type of sensor-system you intend to set up.
4. Make sure the Kinect V2 sensor is plugged in and powered using the Kinect configuration verifier.
5. Build and run the application.
6. If the window displays the Sensor view or a top-down view in which bodies are tracked, the application is correctly receiving data from the sensor.

use instructions

setting up the CTS

1. Firstly it is important to register the sensors and rooms that you wish to surveill. For this, navigate to the “/config”-page on your server.
2. Register any RSS you wish to use by entering its name in the field “Sensor ID” under “Room Surveillance Sensor” and click submit
3. Register any PSS you wish to use by entering its name in the field “Sensor ID” under “Pathway Surveillance Sensor” and click submit
4. Register and Room you wish to surveill by entering the name in the field “Room Number”. Then go into the table of registered sensors and copy the id of the sensors belonging to that room into the Corresponding form fields.
5. Remember that the first Room which a PSS is added to is registered as the sensors “toRoom”, meaning that the direction of the entrance is pointing into that room. This is important so the system correctly adds and subtracts the numbers of inhabitants when they pass between the sensors.
6. Once all rooms and sensors are added, click “reset database counters”. Make sure this is done during a time when all rooms are empty of inhabitants or it might introduce some error.
7. Navigate back to the main page and you should be able to see an overview of the rooms you registered.
8. Once Messages from the connected sensor systems arrive, the numbers of inhabitants will update accordingly and the initially greyed out RSS viewport will turn yellow and display the newest arrived RSS-snapshot along with a timestamp.

seting up the Passageway Sensor System (PSS)

1. Run your build of the application.
2. In the text input field, enter the URL of the sensor endpoint like this: “http://{server-hostname or -IP}:{Port}/sensor/pss/{PSS name}”. An example may look like this: “http://192.168.178.41:3000/sensor/pss/A1entrance”
3. Select a color in the field on the right and draw a line on the ground in the depth-sensor-view. Make sure you are not drawing on any black area so the system can correctly determine the coordinates of the line.
4. Click “Save line”, then “Submit”. If the drawn line is valid, a new window will open.
5. The drawn threshold will be represented as a blue line in the new window. The direction of that threshold will be shown by the red line that is perpendicular to the blue line. Make sure that the direction of the threshold lines up with the direction of the PSS as it is saved in the database, meaning the red line is supposed to point into the direction of the PSSs toRoom. (Remember that the first Room a PSS was assigned to is saved as this Sensors toRoom and the second room is saved as this Sensors fromRoom, see use instructions CTS).
6. If the direction of the Threshold does not line up with the setup, you can flip it by clicking “invert Passage”

seting up the Room Sensor System (RSS)

1. Run your build of the application.
2. In the text input field, enter the URL of the sensor endpoint like this: “http://{server-hostname or -IP}:{Port}/sensor/rss/{RSS name}”. An example may look like this: “http://192.168.178.41:3000/sensor/rss/A1aoi”
3. Specify the interval in which to send the posts to the server. Default of 1000ms is the fastest it allows to update.
4. Click “confirm” to let the application start sending regular updates to the specified URL.

example of use

Tracking of Personell within office buildings to ensure that social distancing measures are being kept.

disclaimer

At this point in time, this system is not reliable enough for any use such as social distancing enforcement. We are not liable for any damages that may occur in actual use.
Before implementing such surveillance systems also consider that privacy is an essential human right that should not be violated.

references

Many parts of this project were based on the repository https://github.com/violetasdev/bodytrackingdepth_course by violetasdev