TL;DR
I added a cronjob to check if the docker daemon is running or not -> if not then restart the service. It works like a champ and I never met the issue again.
* * * * * docker info > /dev/null || systemctl restart docker
Please note that the cronjob should belong to user that has permission to restart docker service, better run it with root user.
Related issue: https://github.com/docker/machine/issues/1458