User-Centric Privacy-by-Design AI-based Adaptive Access Control in a HealthCare Robot for the Elderly

The Industry and Innovation Research Institute (I2Ri) draws on talents, expertise and facilities across Sheffield Hallam University. The vision is to be the leading provider of applied research excellence delivering materials, computing, science and engineering innovations meeting the development needs of industry.

PhD Research Topic

Any AI-based robotic solution needs to guarantee system and data integrity, confidentiality, and service availability to ensure its functionality as intended, otherwise, the system and its data access will not be controlled leading to system and data tampering and manipulation. Care and assistive robots need special attention because they collect, store, and monitor information that is sensitive and private, and they provide services to patients, the elderly, or disabled people related to their health and well-being. Preserving privacy and protecting the user’s data and the care robot’s services becomes a duty of care for the robot developer and service application designers. However, the security solutions should not become a barrier and a burden to elderly users, otherwise, acceptance and adoption will be affected. The chances of human error among elderly users will exponentially grow with age, physical health, and mental condition. So, the security responsibilities laid upon the users should be moved to the robotic system by considering privacy-by-design solutions. Some of the key security responsibilities e.g., authentication, authorization settings, system security, privacy configuration, etc can be managed by the care robot through AI mechanisms to reduce human error. This way, the elderly user need not remember how to authenticate for service access, how to authorize access levels, and conduct accounting for auditing. The aim is to avoid any chances of system’s service disclosure, or unauthorized or unintended data disclosure due to human error during interaction and engaging with the robot.

During this PhD project, you will have the unique opportunity to be under the guidance of distinguished domain experts. These experts come from a variety of specialized fields, including machine learning, cyber security, and robotics. Their combined expertise will provide a comprehensive and multidisciplinary approach to your research, ensuring that you are well-equipped with the necessary knowledge and skills to excel in your project.

You will build a privacy-by-design intelligent access control framework solution that addresses Authentication, Authorization, and Accounting (AAA) using deep learning to control and manage the services and data flow of the care robot in such a way that the care robotic system protects owner’s privacy rights, preferences, and service requirements based on the role of the authorized users when the data pool is unstructured. The research will co-design and experimentally validate a Self-Supervised Robotics System (SSRS) with an Intelligent Data Controller (IDC) deployed in a Robotic Operating System (ROS), which will automatically supervise the human-robot interaction to safeguard users’ privacy, protect user data, and system’s functionality.

The research objectives for this project are listed below:

1. AI-based Seamless Authentication (A): Design and develop a gait-based AI user identification and authentication system in which the user’s gait is actively monitored directly by the robot’s sensor and not supplied by the user or any smart wearables.

2. LLM-based authorization for access (A): This objective will address the access control of the authorized data and services of the care robot using a Large Language Model (LLM), an AI-based solution for an unstructured data pool.

3. Transparent and Secure Accounting (A): The care robotic system must ensure a transparent secure event logging system so that the events and actions are explainable to the user by coupling with a scalable smart contract-based blockchain solution.

4. System Validation of the User-Centric Privacy by Design AAA System: The privacy and secure mechanisms between the robot and user interaction are to be tested and validated with potential stakeholders to confirm acceptance and adoption.

Eligibility

Applicants should hold a 1st or 2:1 Honours degree in a related discipline. A Master’s degree in a related area is desirable. We welcome applications from all candidates irrespective of age, pregnancy and maternity, disability, gender, gender identity, sexual orientation, race, religion or belief, or marital or civil partnership status.

International candidates are required to provide an IELTS certificate with a score of at least 7.0 overall, and a minimum of 6.5 in all components. For further information on English Language requirements, please click here.

For further details on entry requirements, please click here.

How to apply

All applications must be submitted using the online application form. To apply, click here. In your application, be sure to include the title of the project that you are applying for.

As part of your application, please upload:

• A research proposal (max. 1500 words) in your own words, briefly outlining the proposed research, the current knowledge and context referencing key background literature; a proposed methodology or approach to answer the key questions, and any potential significance or impact of the research
• Copy of your highest degree certificate
• Non-UK applicants must submit IELTs results (or equivalent) taken in the last two years and a copy of their passport.

Applicants must provide 2 references, with at least one to be academic. References must be received directly from the referees.

We strongly recommend you contact the lead academic, Jing Wang jing.wang@shu.ac.uk, to discuss your application.

For information on how to apply please visit https://www.shu.ac.uk/research/degrees