Smart City Simulator

High performance virtual environment for testing smart city platforms.

In order to demo the edge4all project, the team designed a virtual environment that simulates a real smart city, represented by large quantities of different devices (in mobility, services they provide, security level, etc...). This simulator is comprised of:

(a) a virtual city, complete with a proper coordinate system and roads;
(b) five personalized, independent, interactive and time-bound services (temperature, noise, air quality, UV Index and water drainage levels) from which devices gather data;
(c) a device management interface, allowing users to add, manipulate and test their own public or secure device in the virtual environment.

Demo links

SCLP: Smart City Lightest Protocol

Lightweight protocol oriented to the security and privacy of IoT devices in Smart Cities

Despite current attempts for achieving lightweight and secure protocols, such as MQTT and CoAP, it is considered that their main aim has not been achieved yet. If a small device intends to send an authenticated message every 30 seconds over MQTT using a x509 certificate, it will use approximately 35MB of data per month*. As this amount of data is currently too expensive with telecommunication operators, specially considering a monthly fee per IoT, this project intends to lower the data consumption to 1MB per month to avoid fees.

To achieve this particular goal, the Smart City Lightest Protocol (SCLP) was developed as a very usecase oriented protocol that adds an extra layer of security to the whole system: the "device fingerprint". This protocol carries a tcp server where the logic is executed and a Redis key-value map where device information is stored. To interact with it, the device must previously register itself with the platform where some parameters are defined (like authentication type, communication frequency, what information will be sent,...). With this setup, the server can "define" the device "personality" (or deviceonality) that will be confirmed every message. If desired upon registration, the server can track the device by status (online of offline), by position (the device location at the last message) and/or by data, keeping an history of the information sent by the device.

Considering the above setup with MQTT (communication with certificate every 30 seconds), with every feature enabled, this new protocol uses 1.08MB per month.

*For this calculation, it is considered that the communication is established every message, because the devices "goes to sleep" to save energy.

Smart Mobility

To analyze, predict and optimize vehicle traffic

This project aims at improving the traffic wherever possible, starting with the cities of Guimarães (ptp. Vitrus) and Braga (ptp. TUB). We are training models with machine learning algorithms that will predict the future traffic so it is possible to improve it and optimize public transportation routes, reducing waiting times. We are also analyzing the data that it's possible to collect in order to analyze it to find patterns and key locations where the traffic is worse than average making it possible to act accordingly in order to make the most of everyone's commuting time.