Description:
The emergence of the Internet of Things (IoT) - the networked connection of objects, processes, data and people - has multiplied the number of
communicating devices (device) in the world, and the problems of interconnection inherent in their networking and accelerated data processing. The
major application sectors of the IoT are in particular industry (smart factory), the city (smart city), and the environment (smart metering and smart
agriculture). There is a broad consensus on the fact that the problem will not only be to manipulate and exploit in the Cloud, all the data emitted by these
objects, but beyond that to manage: i) the network interconnection of thousands of objects under constraints of connection discontinuity, communication
latency and energy consumption for autonomous objects on battery. Currently, 4G/5G technologies appear to be the most promising to meet the
challenges of throughput, latency and connectivity [4], but their mass exploitation in the particular industrial context is not sufficiently proven or tested to
date, and the industrial products are still scarcely available on the market; ii) to ensure the processing of the data generated by the objects, as close as
possible to their source, by the use of on-board analysis and processing technologies, based on the methods of Artificial Intelligence (AI) and machine
learning (ML), using at least Cloud infrastructures, using an edge-computing approach [5]. The constraints related to the reduced capacities of the
microelectronic architectures of communicating objects lead to the need to adapt the methods and algorithms of AI/ML in a light way [6], to the profile
of objects limited in computing capacity and energy.
The expected objectives are both societal and scientific with, among other things, the reduction of energy consumption in a Cloud infrastructure and
implicitly of the carbon footprint of an IoT system, the improvement of connectivity and accessibility of objects, improving decision and action
responsiveness on processes controlled/monitored by communicating objects through IA/ML decision methods and algorithms embedded closer to the
processes.
Thus, the scientific and industrial issues targeted by the thesis work are linked to the design of an autonomous communicating object, supporting
advanced analysis and decision-making capacities, under material and energy constraints, for applications in such varied fields. as smart factory, smart
city, smart metering and smart agriculture.
On the basis of these findings, the research and development work carried out in the thesis under CIFRE agreement will target these 3 objectives:
1. Improve the design of an OKKO communicating object on a hardware and software level to integrate 4G and 5G communication functionalities and
services, in order to meet the challenges of mass deployment with extended connectivity. The minimization of the energy consumption of the object and
the total consumption of the IoT infrastructure will be specification and realization criteria [7];
2. Propose, adapt, develop and experiment with light embedded IA/ML models, allowing to accelerate the analysis by learning and classification of the
data produced by the communicating objects before transmission in the IoT infrastructure, and thus to improve the reactivity to problems. The analysis
must take place at the edge of the network (at the Edge), which avoids the data going back to the central site before being processed.
3. Identify, analyze and integrate cybersecurity aspects into the hardware and software design of communicating objects in relation to the future
European Regulation on the cyber-resilience of electronic equipment (EU Cyber Resilience Act) [8
The experiments may be carried out in the laboratory and on prototype industrial applications operated by OKKO. The thesis work will be carried out
within the framework of a CIFRE convention (Industrial Convention for Training through REsearch) with the OKKO company. The recruited candidate
will carry out his work under the status of research and development engineer within the company and as a doctoral researcher at the CRAN laboratory.
[1] Mekki K, Bajic E., Chaxel F., Meyer F.,., "A comparative study of LPWAN technologies for large-scale IoT deployment", International Journal of
Information & Communication Technology Express, ICT Express, Vol. 5, Issue 1, March 2019, Elsevierhttps://doi.org/10.1016/j.icte.2017.12.005
[2] Mekki K, Bajic E., Chaxel F., Meyer F., "Indoor Positioning System for IoT Device based on BLE Technology and MQTT Protocol", IEEE World
Forum on Internet of Things, Limerick, Ireland, April 2019 ⟨hal-02098503⟩
[3] Projet ANR-21-SOIM-0007-03 Interactive and Intelligent physical assets control system for the Risks Management of hazardous industrial facilities,
CRAN-UMR CNRS 7039, LAMIH-UMR CNRS 8021, société régionale OKKO SaS, spécialisée en conception et développement de solutions pour
l'Internet des Objets
[5] Cao, Keyan & Liu, Yefan & Meng, Gongjie & Sun, Qimeng. (2020). An Overview on Edge Computing Research. IEEE Access. PP. 1-1.
10.1109/ACCESS.2020.2991734.
[6] Bian, Jiang & Arafat, Abdullah Al & Xiong, Haoyi & Li, Jing & Li, Li & Chen, Hongyang & Wang, Jun & Dou, Dejing & Guo, Zhishan. (2022).
Machine Learning in Real-Time Internet of Things (IoT) Systems: A Survey. IEEE Internet of Things Journal. 9. 10.1109/JIOT.2022.3161050.
[7] Mekki K, Bajic E., Meyer F., "Concept and Hardware Considerations for Product-Service System Achievement in Internet of Things". 5th IEEE
International Conference on Wireless Technologies, Embedded and Intelligent Systems, Fez, Morocco, April 2019. {hal-02087881}
[8] Loi européenne sur la cyber-résilience, https://digital-strategy.ec.europa.eu/fr/library/cyber-resilience-act