krasimirvasilev1 / nbu-masters

Serverless Parking Management Cloud-Based System with Microcontroller Endpoints

Providing access management mechanisms by leveraging microcontroller endpoints and Serverless Cloud System and Services.

Linux/Ubuntu local configuration for Lambda and DynamoDB docker container instances

1. Install and configure AWS CLI

apt update && apt upgrade
sudo apt-get install awscli
aws --version
aws configure
	AWS Access Key ID: test
	AWS Secret Access Key: test
	Default region name: eu-central-1

2. Install and configure Docker. Replace {ubuntu-user}!

curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker {ubuntu-user}
sudo systemctl enable docker.service && sudo systemctl enable containerd.service

HINT - You can check your current user by executing whoami in the console!

3. Define and run Docker container instance on which the dynamo-db instance will run. The orchestration is managed by docker-compose.yaml file which specify what image is required.

cd dynamodb_local && mkdir -p ./docker/dynamodb && docker-compose up -d

4. Give required permissions. Create database table and populate it with content.

cd ../utils && bash dynamo_local_setup.sh

5. Open lambda config file lambdas/check_plate/app/app.py. Replace IP-ADDRESS placeholder with the IPv4 address of your machine !

dynamodb = boto3.resource('dynamodb',
                          aws_access_key_id="test",
                          aws_secret_access_key="test",
                          region_name="eu-central-1",
                          endpoint_url="http://{IP-ADDRESS}:8000")

HINT - You can check your IP address by executing : Linux - ip a Windows - ipconfig | findstr /R /C:"IPv4 Address"

6. Build and run Lambda. Install needed dependencies, setup build environment and compile environment for lambda function. Install prerequisites. Wait till the process is finished.

cd ../lambdas/check_plate && docker build -t localfunction:latest . 
docker run -d -p 9000:8080 localfunction:latest
sleep 10

7. Test cloud serverless functionality. Please replace IP-ADDRESS placeholder with the IPv4 address of your machine !

cd ../../utils/images

(echo -n '{"image": "'; base64 P00399.jpeg; echo '"}') | curl -sH "Content-Type: application/json" -d @- "http://{IP-ADDRESS}:9000/2015-03-31/functions/function/invocations" >> ../results.log

**8. Check the result in /utils/results.log. **

Status code 200 in that case means that plate HH0L7687 is recognized by the OpenALPR software and exists in the DynamoDB table and accordingly receives a green light to pass from our access control system.

Raspberry Pi setup

1. Insert SD card into the Raspberry Pi and install the OS as per https://www.raspberrypi.org/documentation/installation/installing-images/

2. Once the OS is installed, execute the following commands. Wait until all prerequisite packages are installed !

apt update && apt upgrade && apt install -y jq motion

**3. Install Docker by executing the following commands:

curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker pi
sudo systemctl enable docker.service && sudo systemctl enable containerd.service

4. Apply the configuration file used by Motion. This can be done with the below command executed from the root directory of the repository. Placeholder should be replaced regarding the IP address of the machine that is running.

scp ./raspberry/motion.conf pi@{RASPBERRY_IP}:/etc/motion

5. Wire the LEDs and the resistors following this schematic - https://images.app.goo.gl/yhGTY9iL2KfGztLc7

6. Open rasp_check config file raspberry/rasp_check.sh. Replace IP-ADDRESS placeholder with your IPv4 addresss! You should provide the local IP of the host where Lambda lives!

CLOUD_URL=http://{IP-ADDRESS}:9000/2015-03-31/functions/function/invocations

HINT - You can check your IP address by executing : ip a OR ipconfig | findstr /R /C:"IPv4 Address"

7. Add the rasp_check.sh file to /var/lib/motion by executing the following command from the root directory of the repository.

scp ./raspberry/rasp_check.sh pi@{RASPBERRY_IP}:/var/lib/motion

8. Еnable the digital output of Raspberry PI. That's the way of controlling the voltage applied to the diodes, which determines whether they will light up or not.

cd /sys/class/gpio
echo "17" > ./export
echo "out" > ./gpio17/direction
echo "18" > ./export
echo "out" > ./gpio18/direction

echo "1" > /sys/class/gpio/gpio18/value

9. Start the motion service

sudo systemctl start motion

10. Pull docker image which contains OpenALPR software for automatic number-plate recognition. The image will be used for an offline check or in other words when there is no internet access.

docker pull krasimirvasilev1/nbu-alpr