Securing the internet of robotic things: a comprehensive review on machine learning-based intrusion detection
Matilda Nkoom, Sena Hounsinou, Garth V. Crosby
Abstract
The Internet of Robotic Things (IoRT) emerges from integrating robotics into the Internet of things (IoT). The dynamic nature of the IoRT environment poses unique security challenges. Implementing adequate security measures can be challenging due to the limited resources, real-time operational requirements, and heterogeneous nature of the mobile network nodes. Machine learning-based intrusion detection systems (IDSs) have significantly improved the security of IoT devices, and other Cyber-Physical Systems (CPSs) by providing embedded intelligence to mitigate cyber threats. Although machine learning models have been extensively used in intrusion detection for IoT systems, their specific application in the context of IoRT remains significantly unexplored. As a result, this paper proposes implementing machine learning-based IDSs to enhance the security of IoRT systems. To provide a comprehensive overview of this emerging research domain, this paper presents a comparative review to differentiate between CPSs, IoT, and IoRT to establish a basis for extending existing machine learning-based IDSs and, the need for more advanced solutions. It discusses the challenges in implementing these security solutions and outlines future research directions. Additionally, the paper introduces an initial concept of a privacy-preserving security framework for IoRT systems.