PhD Privacy-Preserving Fault Detection in Cyber-Physical Systems
PhD Privacy-Preserving Fault Detection in Cyber-Physical Systems
You cannot apply for this job anymore (deadline was 31 Aug 2020).
Browse the current job offers or choose an item in the top navigation above.
PhD Privacy-Preserving Fault Detection in Cyber-Physical Systems (V35.4492) in the group Dynamics & Control.
Job description
The Dynamics and Control (D&C) group at Eindhoven University of Technology (TU/e) is looking for a talented PhD candidate to work on developing analysis and synthesis tools for privacy-preserving fault detection schemes for networked Cyber-Physical Systems (CPS).
Technology companies today manufacture and sell high-tech equipment capable of measuring, processing, and transmitting operational data in real-time over the cloud. Companies collect data of the day-to-day operation of equipment to monitor the life and proper operation of their products. Online equipment monitoring allows them to forecast failures and schedule maintenance before a critical fault (event) occurs. However, sharing operational data might not be attractive to end-users as private/sensitive information of the products (services) they manufacture (provide) could be disclosed. There are two potential privacy threats in this setting: 1) the customer does not wish the manufacturer (service provider) to infer certain information (e.g., manufactured products specs); and 2) it is the communication channel itself (e.g., the internet), what the customer does not trust. Conversely, online equipment monitoring extends the life of equipment, decreases production bottlenecks, and prevents high and unexpected costs. That is, we have two opposite aspects here, customers would like to share data to improve performance but they do not want to share it for privacy reasons. It is therefore attractive for both parties (customers and manufacturers) to have masking mechanisms that allow for: 1) coding operational data before disclosure so that private information is hidden (in some appropriate sense); and 2) the coded data to be used by the manufacturer (service provider) to detect faults and schedule maintenance.
The broad goal of this project is to develop fundamental systems, control, and information theoretic tools that allow constructing privacy-preserving fault detection schemes for networked Cyber-Physical Systems. We aim at synthesizing real-time fault detection algorithms that, on the one hand, satisfy the required prescribed detection performance; and, on the other hand, guarantee a private exchange of system data over (potentially untrusty) communication networks. The main research questions to be addressed in the project are: Depending on the class of systems under study (e.g., linear, nonlinear, stochastic, hybrid, etc.), how to properly select privacy metrics that make sense from the point of view of dynamical systems? Given a particular privacy metric and a fault detection scheme, how to quantify the privacy level (information leakage) provided by the detection scheme? And, what synthesis tools can be used to systematically design coding functions and fault detection algorithms to maximize privacy and guarantee prescribed detection performance?
The Dynamics and Control (D&C) group at TU/e trains the next-generation of students to understand and predict the dynamics of complex engineering systems in order to develop advanced control, estimation, planning, and learning strategies which are at the core of the intelligent autonomous systems of the future. Autonomous vehicles, fully automated industrial value chains, high-tech systems, collaborative robots in unstructured environments, intelligent medical devices, automated transportation networks, soft robotics, together with sustainable automotive technology are key examples of the broad application domain of the (D&C) group. The design of these systems requires a thorough understanding of their underlying dynamics. Therefore, the first focal point of our research is on both data-based and first-principles-based modelling, model complexity management, and dynamic analysis of complex, multi-physics and multi-disciplinary engineering systems. Building on this foundation, our second focal point is on 'making autonomous systems smarter'. To this end, we develop both model- and data-based sensing, planning, and learning and control technologies to provide autonomous systems with the intelligence needed to guarantee performance, robustness, and safety. Combining the investigation on both dynamics and control theory in one section allows to take on these challenges standing in a privileged position. In particular, it enables us to educate uniquely skilled engineers and researchers as well as to valorize our research together with the high-tech, automotive and energy sectors.
Brainport Eindhoven Region
The Brainport Eindhoven is a world-class top technology region, in which companies, governments, and educational institutions (the so-called triple helix collaboration) work together on advancing technology for humanity. Brainport Eindhoven is among Europe's most prominent and innovative high-tech centers, where high-tech and design are combined with high-end manufacturing industry and entrepreneurship. Geographically situated in the southern part of the Netherlands, the Brainport region has a workforce of 400,000 people from all over the world working on high-tech solutions in areas such as health, mobility, energy, and nutrition. The region generates, by far, the most patents per thousand inhabitants in the world.
Technology companies today manufacture and sell high-tech equipment capable of measuring, processing, and transmitting operational data in real-time over the cloud. Companies collect data of the day-to-day operation of equipment to monitor the life and proper operation of their products. Online equipment monitoring allows them to forecast failures and schedule maintenance before a critical fault (event) occurs. However, sharing operational data might not be attractive to end-users as private/sensitive information of the products (services) they manufacture (provide) could be disclosed. There are two potential privacy threats in this setting: 1) the customer does not wish the manufacturer (service provider) to infer certain information (e.g., manufactured products specs); and 2) it is the communication channel itself (e.g., the internet), what the customer does not trust. Conversely, online equipment monitoring extends the life of equipment, decreases production bottlenecks, and prevents high and unexpected costs. That is, we have two opposite aspects here, customers would like to share data to improve performance but they do not want to share it for privacy reasons. It is therefore attractive for both parties (customers and manufacturers) to have masking mechanisms that allow for: 1) coding operational data before disclosure so that private information is hidden (in some appropriate sense); and 2) the coded data to be used by the manufacturer (service provider) to detect faults and schedule maintenance.
The broad goal of this project is to develop fundamental systems, control, and information theoretic tools that allow constructing privacy-preserving fault detection schemes for networked Cyber-Physical Systems. We aim at synthesizing real-time fault detection algorithms that, on the one hand, satisfy the required prescribed detection performance; and, on the other hand, guarantee a private exchange of system data over (potentially untrusty) communication networks. The main research questions to be addressed in the project are: Depending on the class of systems under study (e.g., linear, nonlinear, stochastic, hybrid, etc.), how to properly select privacy metrics that make sense from the point of view of dynamical systems? Given a particular privacy metric and a fault detection scheme, how to quantify the privacy level (information leakage) provided by the detection scheme? And, what synthesis tools can be used to systematically design coding functions and fault detection algorithms to maximize privacy and guarantee prescribed detection performance?
The Dynamics and Control (D&C) group at TU/e trains the next-generation of students to understand and predict the dynamics of complex engineering systems in order to develop advanced control, estimation, planning, and learning strategies which are at the core of the intelligent autonomous systems of the future. Autonomous vehicles, fully automated industrial value chains, high-tech systems, collaborative robots in unstructured environments, intelligent medical devices, automated transportation networks, soft robotics, together with sustainable automotive technology are key examples of the broad application domain of the (D&C) group. The design of these systems requires a thorough understanding of their underlying dynamics. Therefore, the first focal point of our research is on both data-based and first-principles-based modelling, model complexity management, and dynamic analysis of complex, multi-physics and multi-disciplinary engineering systems. Building on this foundation, our second focal point is on 'making autonomous systems smarter'. To this end, we develop both model- and data-based sensing, planning, and learning and control technologies to provide autonomous systems with the intelligence needed to guarantee performance, robustness, and safety. Combining the investigation on both dynamics and control theory in one section allows to take on these challenges standing in a privileged position. In particular, it enables us to educate uniquely skilled engineers and researchers as well as to valorize our research together with the high-tech, automotive and energy sectors.
Brainport Eindhoven Region
The Brainport Eindhoven is a world-class top technology region, in which companies, governments, and educational institutions (the so-called triple helix collaboration) work together on advancing technology for humanity. Brainport Eindhoven is among Europe's most prominent and innovative high-tech centers, where high-tech and design are combined with high-end manufacturing industry and entrepreneurship. Geographically situated in the southern part of the Netherlands, the Brainport region has a workforce of 400,000 people from all over the world working on high-tech solutions in areas such as health, mobility, energy, and nutrition. The region generates, by far, the most patents per thousand inhabitants in the world.
Specifications
- max. 38 hours per week
- Eindhoven View on Google Maps
:fill(white)/logos/tue-en-wide_acOLN2b.png)
Eindhoven University of Technology (TU/e)
Requirements
We are looking for a recently graduated, talented, and enthusiastic candidate who meets the following criteria:
- Master of Science degree (or an equivalent university degree) in Mechanical or Electrical Engineering, Computer Science, Applied Physics, Applied Mathematics, Robotics, or related disciplines.
- Strong analytical skills.
- Strong Matlab programming skills.
- Fluent in spoken and written English.
- Experience in model-based fault detection and control schemes for mechanical systems is a plus.
Conditions of employment
We offer:
- A meaningful job in a dynamic and ambitious university with the possibility to present your work at international conferences.
- A full-time employment for four years, with an intermediate evaluation after one year.
- To support you during your PhD and to prepare you for the rest of your career, you will have free access to a personal development program for PhD students (PROOF program).
- A gross monthly salary and benefits in accordance with the Collective Labor Agreement for Dutch Universities. A salary is offered starting at € 2395 per month (gross) in the first year, increasing up to € 3061 per month (gross) in the last year.
- Additionally, an annual holiday allowance of 8% of the yearly salary, plus a year-end allowance of 8.3% of the annual salary.
- A broad package of fringe benefits, including an excellent technical infrastructure, moving expenses, and savings schemes.
- Family-friendly initiatives are in place, such as an international spouse program, and excellent on-campus children day care and sports facilities.
