This is an excellent opportunity to work in the exciting area of Artificial Intelligence of Things (AIoT), focusing on developing new security and reliability mechanisms, which can be employed within several industrial sectors such as automotive, aeronautics, medical and smart infrastructures. The Loughborough University research team has been vigorously communicating research activities to a broad audience through journal papers and invited talks in international industrial and academic conferences.
Artificial intelligence (AI) and machine learning (ML) models are being incorporated in resource-constrained Internet of Things (IoT) devices, which typically rely on reduced memory footprint and low-performance processors. Software libraries and application programming interfaces (APIs) have been proposed to enable the execution of such models in the underlying devices. Such libraries/APIs are devoted to streamlining the design and development of embedded deep learning-based applications through the fine-tuning of pre-trained network models, thus enabling their efficient execution in edge-computing platforms. For the time being, available APIs focus on optimising such models considering their accuracy and performance over a given dataset.
With the growing adoption of AI/ML models in safety-critical embedded systems (e.g., medical devices, autonomous vehicles) increases the demand for safe and reliable models that are immune to external threats. To reach safety and reliability levels comparable to those required by high industrial standards, it is imperative to supply edge-computing platforms with appropriate cost-efficient mechanisms that comply with strict safety and reliability requirements (e.g., end-to-end security and privacy) while reducing resource usage.
This PhD research aims to investigate novel, cost-effective techniques and software libraries that can improve the reliability and security of the emerging Artificial Intelligence of Things (AIoT) systems.
Show more