This article was edited by SPRITE+ Research Associate Dmitry Dereshev, with responses and edits from Lecturer of Systems Security Soteris Demetriou.
The spotlight today is on Soteris Demetriou – Lecturer of Systems Security at the Department of Computing at Imperial College London, Director of the Applications, Platforms, and Systems Security Research Lab, and a SPRITE+ member. Some of Soteris’ latest publications include:
How would you describe your job to a 12-year-old?
I'd like to think of my role as a detective. Detectives try to identify the people behind crimes by analysing human behaviour and physical crime scenes. I'm interested in identifying dangerous actions performed in the digital world by analysing the behaviour of computer programs. For example, I am trying to understand if and how mobile apps in smartphones can spy on you or even steal money.
Sometimes I must think like an attacker to discover how such dangerous actions can be performed by those programs before they do. When I figure that out, I work on changing the way digital technologies (e.g. smartphones) operate to make sure such actions never happen.
Could you describe what you do during a typical workday?
As most other academics I wear multiple hats throughout the day. I teach, research, do administrative work, and manage people. It's exciting, I never get bored.
A typical day would involve either delivering a lecture or performing another teaching-related activity such as developing materials or marking assignments and projects. Then, I might attend several meetings where we discuss research projects. This is my favourite part of the day. Nothing surpasses the excitement of having one of those eureka! moments, when an idea for solving a problem strikes us. Then, I would typically work on reading, reviewing, or writing an academic paper or grant proposal. Depending on the day, I might also have to help with administrative tasks, for example reviewing applications and interviewing candidates for student or faculty admissions.
Could you describe a challenging project that you’ve recently worked on?
In a recent collaboration with Hewlett-Packard Labs, we designed a novel technique for improving outdoor vehicle positioning.
Have you ever had your Maps app show your car in the wrong lane or even the wrong road? To figure out where you are, the app does 2 things: 1. it constantly check built-in sensors to predict where your car should be on the map, and which direction it moves; 2. every now and then it checks with GPS to correct for any errors. The GPS part is the problem. Tall buildings on the sides of the road can disrupt the GPS signal, and the accuracy drops to 30-50 meters instead of the usual 5.
To better predict where a car is on the road, we came up with an inter-vehicle approach where you can use better sensors (like cameras and sonars) in modern cars to help older cars find where they are. With this approach (called CoDrive) we could achieve a 90% reduction in cumulative GPS error for older cars, and a 30% error reduction in sensor-rich cars.
This was a very challenging project but also very fun and rewarding. To test our techniques in practice, we had to create a realistic traffic environment. We rented several cars and blocked a road segment within the Hewlett-Packard campus in Palo Alto for our experiments.
The most challenging part was emulating a sensor-rich vehicle. We had to develop our own car mount which we attached to the roof of a vehicle, carrying a depth sensor and four mobile phones livestreaming time, location, depth, and visual information from the vehicle’s surroundings. We also had to develop a way for in-car computer to make use of that data and display useful information to the driver. With all that we could reliably detect vehicles and estimate their distance from the emulated sensor-rich car.
This work relies heavily on good data from modern sensors. Those sensors can be fooled, however, and when this is done on purpose, the car can abruptly stop and cause an accident. We are now developing techniques for detecting when a vehicle is under such attacks, so that a car can distinguish real obstacles in front of it from fakes in an efficient, reliable, and robust manner.
What training/experience did you have at the start of your career?
I completed a 5-year degree in Electrical and Computer Engineering (University of Patras, Greece) and then an MSc and a PhD degree in Computer Science (University of Illinois at Urbana-Champaign, USA) focusing on mobile operating systems security. When I decided to pursue an MSc, I knew very little about research, and my skills for undertaking rigorous and independent research projects were underdeveloped. My research skills were gradually improved throughout my postgraduate degrees. I'm still improving these skills and I anticipate it to be a lifelong process.
Tasks that helped me grow during my postgraduate degree include reviewing papers for conferences and journals, delivering lectures, preparing and marking assignments, and attending conferences to present my work. Learning from and with others is also important in developing expertise.
I was fortunate in my early postgraduate degree days to have been working closely with more senior PhD students and world-renowned academics such as my advisor Professor Carl A. Gunter and Professor XiaoFeng Wang.
During my first year as a Lecturer I have also attended various workshops in my host institution which helped me work on supervision, tutoring and teaching skills but also get familiarized with my host institution's expectations and regulations. Coming from the US research scene I also had to adapt to the UK and EU systems. I have been trying to exploit various opportunities for attending industry and academic events relevant to my research interests throughout the UK and Europe.
How did you get into your current role?
During the last year of my PhD I was struggling to choose whether I would like to pursue a career at an industry research lab or a tenure track career in academia. I prepared and applied for both roles and was fortunate to receive tempting offers in both directions. This made my decision even harder. At the end I chose the academic path since I not only like to be constantly challenged but I also greatly value the flexibility and intellectual freedom that comes with an academic environment.
What do you wish you'd known when you started your career?
Firstly, I wish I had understood the importance but also the limitations of the peer review process. Academics disagree often and while receiving multiple reviews should eliminate variance, you are highly likely to receive reviews with a dismissive and sometimes even aggressive tone. Understandably at the beginning it might be challenging to set aside emotions, but it is important to do so, and to focus instead on the gist of the review which should hopefully help improve your work. Unfortunately, it is also likely for some reviews to contain no useful information, but this is rooted in a deeper issue related to our responsibility in training our students on how to give constructive feedback.
Secondly, I wish I had recognized the importance of developing professional relationships earlier. Even though I was fortunate to work with numerous people, I would also consciously try to get involved with more professional activities such as student leadership programs, poster sessions, student conference committees. This is useful not only for engaging in an academic discourse but also for forging relationships that can develop into fruitful long-term collaborations.
And thirdly, the importance of maintaining a good balance between work and personal life as this is paramount for one’s mental wellbeing. Academia is both a very competitive and gratifying environment. On one hand it requires hard work to keep up and lead in research. On the other hand, we choose the problems to work on according to our interests and thus it is easy to spend disproportionate amount on time on them. It is important to keep the big picture in mind and afford time for other things in life that make you happy, whether that is spending time with family and loved ones, travelling, attending social events or anything else one might value.
What would you recommend to people who want to follow in your footsteps?
Going beyond strengthening one's background through university classes, I would recommend students to get involved with research activities as early as possible. Getting exposed to rigorous methodologies, reviewing and discussing academic papers, helping in writing a research article – all contribute to developing fundamental skills and generating evidence of research maturity and potential, which can help with being admitted to a doctorate program.
Then, if you do end up pursuing a PhD, work with your supervisor to select a good topic and consciously avoid distractions that drive you away from it. As Professor Jeff Erickson once told me: “a good topic lies at the intersection of what you enjoy or value, what you do well, and what people will pay for”. If you are like me, you often get excited by new topics and problems. However, especially at the beginning, it is important to develop expertise on a well identified subject, technique or application domain.
I also encourage people to never despair. Academia can be a harsh place where we are constantly being criticized and a lot of junior PhD candidates end up suffering from impostor syndrome. Sometimes, research results can take months, even years to get published while peer review feedback is not always constructive. Sometimes it might feel like everything has already been done and there is little left to explore. Sometimes it might feel like the results are bad, or that everyone else’s work is better than yours.
You should try to focus on yourself and avoid comparing yourself to others, while working on “good” topics at your own pace. At the end of the day, if you like what you do, you will be happy, and an academic career is one of the few environments that affords you the flexibility to work on what you like, with who you like, when you like.
What troubles did you have progressing through your career?
Everything I mentioned above did happen throughout my career to some degree. I remember feeling very disappointed after my first paper reviewers rejected my hard work. However, throughout academia you learn how to manage rejection, risk, and successes very well, and how to use them to develop both personally and professionally.
At the current stage of my career, I think the most challenging aspect is securing funding. Sponsors want to reduce risk in their investments and therefore look for evidence in the track record of the applicant that ensure success of the proposed project. Early career researchers can rarely compete with track records of more senior academics that have already successfully delivered high-risk projects. That's why I believe in the importance of growing your network which will allow you to establish collaborations with more experienced academics to jump-start your career. Networks such as SPRITE+ have the potential to become instrumental in this process.
What one stereotype would like to dispel about your job or industry?
Most people think of academics as teachers, or as introverts locked in their basements performing experiments in isolation. This is not true. Academics usually develop very strong social, communication and leadership skills as these are fundamental for their success. Teaching is an important aspect of the role, but only constitutes a small fraction of it.
Also, reading and writing are common activities which require focus. However, academics also spend a significant amount of time meeting people to discuss research projects and proposals, to define management and administrative operations for their Lab, Department or College, conducting interviews for hiring and growing their Lab, and attending conferences and workshops, sometimes in Hawaii 😉.
How would you describe your research or business interest in relation to SPRITE+?
SPRITE+ is a network which aims to increase engagement from both academia and industry, realized in multidisciplinary relationships and projects related to trust, identity, privacy and security. My research interests are in agreement with SPRITE+ goals since my work focuses on the security and privacy of mobile and cyber-physical systems.
How do you hope to benefit from working with SPRITE+ network?
I would like to take advantage of SPRITE+ network to introduce myself to the UK cybersecurity community and establish fruitful national collaborations with academics and other research stakeholders with similar interests. I want to draw on expertise from SPRITE+’s interdisciplinary pool of skills to deliver stronger research outcomes. The areas of focus of SPRITE+ closely align with my research interests and I believe I would be able to propose highly relevant projects on exciting new directions at the intersection of mobile and Internet of Things (IoT) systems and cybersecurity.
Which of the SPRITE+ Challenge Themes can you relate to from the job that you do? How does it impact your role?
My work mostly carries similarities with the Digital Vulnerabilities theme. In our Lab we are analysing digital systems, aiming to identify and mitigate security vulnerabilities which an attacker can take advantage of. We are also working to understand and build tools to detect and prevent digital information leakage which can result in privacy violations.
Other Themes are also relevant to our Lab's work as they are concerned with similar challenges. For example, the Accountability and Ethics in a Digital Ecosystem theme is concerned with identifying how we can enable privacy- and accountability-by-design in digital technologies. A lot of our work revolves around designing systems and protocols which are both secure and privacy-preserving.
Applications are open to individuals from academia and professional practice (non-academic) to attend an online sandpit on Digital Vulnerabilities in July 2021. Up to £160k of SPRITE+ funding will be made available to fund interdisciplinary projects.