Mycelium 4.0

The Mycelium Project

The Mycelium 4.0 project is part of a collaboration between the National Institute of Applied Sciences in Rennes (INSA) and the Observatory of Universe Sciences in Rennes (OSUR). Its main objective is to establish an infrastructure for processing data produced by geo-distributed sensors, to better understand environments and their evolution, particularly in response to human activities.

These sensors allow tracking and analyzing ongoing environmental changes to develop appropriate solutions. The Beaulieu university campus in Rennes (La Croix-Verte) serves as the study framework for our project.

La Croix Verte

Currently in the process of renaturalization, this green space provides a place for nature in the heart of the Beaulieu campus.

The Team

Students

Amine LAHMAMSI

Arno LECRIVAIN

Elise BOTTOIS

Thibault DUFOURCQ

Mohamed ALLAY

Supervisors

Nikolaos PARLAVANTZAS

Volodia PAROL-GUARINO

The Architecture

An Architecture Based on Fog Computing

Our infrastructure is part of the Fog Computing paradigm, a distributed computing architecture that extends cloud computing to the network edges. Fog Computing allows data to be processed closer to its generation source, thus reducing latency and bandwidth consumption.

Our infrastructure includes sensors, a LoRaWAN gateway, and a cluster composed of five Raspberry Pis.

Located at La Croix-Verte, the sensor operates continuously to collect environmental data. These data are then sent to the gateway via LoRaWAN, at a predefined frequency.

The gateway and the cluster are installed in the computer building of INSA Rennes.

The data collected by the gateway are transmitted to the cluster via an Ethernet connection.

The cluster ensures data processing.

For processing requiring more resources, a VPS takes over. When the load on the cluster becomes too high, it delegates certain tasks to the VPS, in exchange for a longer communication delay and higher bandwidth consumption.

Components of Our Architecture

Sensor

A device that measures environmental parameters such as temperature, humidity, etc.

RaspberryPi

A mini-computer used for processing data collected by sensors.

RaspberryPi Cluster

A set of interconnected Raspberry Pis, where sensor data is processed and analyzed in a distributed manner to perform all scenario-related processing.

Gateway

A device that transmits sensor data to the Raspberry Pi cluster via an Ethernet connection.

VPS

A virtual server hosted at INSA used to perform calculations when the Raspberry Pi cluster's capacity is exceeded, or to manage services and absorb surplus load

New Scenarios and Additions to Mycelium 4.0

To improve the autonomy of the SoLo node and optimize the processing of data received by the Raspberry Pi cluster, new scenarios and architecture optimization features have been added, in addition to the existing scenarios in Mycelium 3.0/2.0 versions.

Frequency Change Scenario

The frequency change scenario consists of dynamically adapting the sampling frequency of sensors based on environmental conditions and detected events. This approach allows reducing energy consumption while maximizing the relevance of collected data. The system reacts in real-time to environmental variations, adjusting the frequency according to the results of statistical analyses and previously detected climate events, thus ensuring optimal resource management and effective environmental monitoring.

Flood Scenario

The main objective of this scenario is to analyze the correlation between several key environmental variables, such as rainfall amount, water course height, and groundwater depth. By cross-referencing these data, the challenge is to identify significant patterns and relationships to predict a potential imminent flood risk. This prediction will be made using predictive algorithms, thus allowing anticipation of critical periods where water accumulation, caused by excessive rainfall or rising water levels, could lead to overflows.

Proxy Implementation within the Architecture

Integrating a proxy into the architecture allows optimal and automatic resource management by regulating message distribution between the Raspberry Pi cluster and the VPS. For example, if the VPS is connected and has sufficient resources, the proxy delegates complex predictive calculations to the VPS.

Features

Intuitive Dashboard

Thanks to Grafana, a clear and intuitive user interface allows real-time visualization of environmental data.

Energy Efficient

Architecture optimized for minimal energy consumption.

Real-Time Alerts

Notification system to alert in case of critical situations.

Scalability

The system can autonomously transfer part of the processing from the cluster to the VPS based on the workload.