6:00 p.m.
Smile Grenoble
Pierre Ficheux
Smile is organizing a meetup on the theme of embedded systems, energy and environment.
Energy and the environment have never been so present in the news. The summer of 2022 was exceptionally hot with maximums reaching 40°C very regularly.
As a result, 4 major fires ravaged more than 35,000 ha of forest in Gironde. Embedded systems and IoT are today technologies that can be used to limit environmental degradation. After some generalities on energy and environmental subjects, we will see how the optimization of systems can allow the implementation of very low consumption sensors useful for environmental protection (fire detection).
Energy and the environment have never been so present in the news. The summer of 2022 was exceptionally hot with maximums reaching 40°C very regularly.
As a result, 4 major fires ravaged more than 35,000 ha of forest in Gironde.
For several years, numerous publications have mentioned the energy footprint of “data centers” but embedded systems and other IoT sensors and gateways (less known to the general public) are rarely mentioned.
However, due to the number effect, the impact of these systems is not negligible, also knowing that many homes have (at least) a high-speed connection and therefore several pieces of equipment which most often operate 24/7.
The evolution of connectivity towards personal vehicles (electric and/or connected cars) has a significant impact both on an energy and environmental level through the consumption of data (therefore access to data centers) and the use of components storage which have a limited lifespan inversely proportional to the volume of data written. These components must be replaced, and therefore produced, with consequences on GHG (CO2) emissions. However, we will see how certain manufacturers use a “low tech” approach to limit these emissions.
We will also see how the choice of languages and tools can influence the energy footprint. We will give some examples based on increasingly used software components, such as Android Automotive OS (based on a Linux kernel).
We will discuss the techniques used for areas in which energy has always been scarce (isolated systems such as space applications or autonomous “IoT” sensors).
We will also present a POC from the Vigi Fire project whose aim is to detect fire outbreaks using a low-consumption sensor and an IoT architecture. To do this, we will discuss the use of the “deep sleep” operating mode which limits consumption to around 200 µA on ESP32. We will present concrete comparative results as part of the study of a low consumption sensor (several months or even years of autonomy).