Marijn Janssen

Dewant Katare, Diego Perino, Jari Nurmi et al.

Autonomous driving services rely heavily on sensors such as cameras, LiDAR, radar, and communication modules. A common practice of processing the sensed data is using a high-performance computing unit placed inside the vehicle, which deploys AI models and algorithms to act as the brain or administrator of the vehicle. The vehicular data generated from average hours of driving can be up to 20 Terabytes depending on the data rate and specification of the sensors. Given the scale and fast growth of services for autonomous driving, it is essential to improve the overall energy and environmental efficiency, especially in the trend towards vehicular electrification (e.g., battery-powered). Although the areas have seen significant advancements in sensor technologies, wireless communications, computing and AI/ML algorithms, the challenge still exists in how to apply and integrate those technology innovations to achieve energy efficiency. This survey reviews and compares the connected vehicular applications, vehicular communications, approximation and Edge AI techniques. The focus is on energy efficiency by covering newly proposed approximation and enabling frameworks. To the best of our knowledge, this survey is the first to review the latest approximate Edge AI frameworks and publicly available datasets in energy-efficient autonomous driving. The insights and vision from this survey can be beneficial for the collaborative driving service development on low-power and memory-constrained systems and also for the energy optimization of autonomous vehicles.

Prachi Bagave, Marcus Westberg, Marijn Janssen et al.

AI is transforming the healthcare domain and is increasingly helping practitioners to make health-related decisions. Therefore, accountability becomes a crucial concern for critical AI-driven decisions. Although regulatory bodies, such as the EU commission, provide guidelines, they are highlevel and focus on the ''what'' that should be done and less on the ''how'', creating a knowledge gap for actors. Through an extensive analysis, we found that the term accountability is perceived and dealt with in many different ways, depending on the actor's expertise and domain of work. With increasing concerns about AI accountability issues and the ambiguity around this term, this paper bridges the gap between the ''what'' and ''how'' of AI accountability, specifically for AI systems in healthcare. We do this by analysing the concept of accountability, formulating an accountability framework, and providing a three-tier structure for handling various accountability mechanisms. Our accountability framework positions the regulations of healthcare AI systems and the mechanisms adopted by the actors under a consistent accountability regime. Moreover, the three-tier structure guides the actors of the healthcare AI system to categorise the mechanisms based on their conduct. Through our framework, we advocate that decision-making in healthcare AI holds shared dependencies, where accountability should be dealt with jointly and should foster collaborations. We highlight the role of explainability in instigating communication and information sharing between the actors to further facilitate the collaborative process.

Dewant Katare, David Solans Noguero, Souneil Park et al.

The accuracy and fairness of perception systems in autonomous driving are essential, especially for vulnerable road users such as cyclists, pedestrians, and motorcyclists who face significant risks in urban driving environments. While mainstream research primarily enhances class performance metrics, the hidden traits of bias inheritance in the AI models, class imbalances and disparities within the datasets are often overlooked. Our research addresses these issues by investigating class imbalances among vulnerable road users, with a focus on analyzing class distribution, evaluating performance, and assessing bias impact. Utilizing popular CNN models and Vision Transformers (ViTs) with the nuScenes dataset, our performance evaluation indicates detection disparities for underrepresented classes. Compared to related work, we focus on metric-specific and Cost-Sensitive learning for model optimization and bias mitigation, which includes data augmentation and resampling. Using the proposed mitigation approaches, we see improvement in IoU(\%) and NDS(\%) metrics from 71.3 to 75.6 and 80.6 to 83.7 for the CNN model. Similarly, for ViT, we observe improvement in IoU and NDS metrics from 74.9 to 79.2 and 83.8 to 87.1. This research contributes to developing reliable models while enhancing inclusiveness for minority classes in datasets.

Mannat Kaur, Michel van Eeten, Marijn Janssen et al.

Instead of only considering technology, computer security research now strives to also take into account the human factor by studying regular users and, to a lesser extent, experts like operators and developers of systems. We focus our analysis on the research on the crucial population of experts, whose human errors can impact many systems at once, and compare it to research on regular users. To understand how far we advanced in the area of human factors, how the field can further mature, and to provide a point of reference for researchers new to this field, we analyzed the past decade of human factors research in security and privacy, identifying 557 relevant publications. Of these, we found 48 publications focused on expert users and analyzed all in depth. For additional insights, we compare them to a stratified sample of 48 end-user studies. In this paper we investigate: (i) The perspective on human factors, and how we can learn from safety science (ii) How and who are the participants recruited, and how this -- as we find -- creates a western-centric perspective (iii) Research objectives, and how to align these with the chosen research methods (iv) How theories can be used to increase rigor in the communities scientific work, including limitations to the use of Grounded Theory, which is often incompletely applied (v) How researchers handle ethical implications, and what we can do to account for them more consistently Although our literature review has limitations, new insights were revealed and avenues for further research identified.

Aaron Yi Ding, Gianluca Limon De Jesus, Marijn Janssen

The security of the Internet of Things (IoT) has attracted much attention due to the growing number of IoT-oriented security incidents. IoT hardware and software security vulnerabilities are exploited affecting many companies and persons. Since the causes of vulnerabilities go beyond pure technical measures, there is a pressing demand nowadays to demystify IoT "security complex" and develop practical guidelines for both companies, consumers, and regulators. In this paper, we present an initial study targeting an unexplored sphere in IoT by illuminating the potential of crowdsource ethical hacking approaches for enhancing IoT vulnerability management. We focus on Bug Bounty Programs (BBP) and Responsible Disclosure (RD), which stimulate hackers to report vulnerability in exchange for monetary rewards. We carried out a qualitative investigation supported by literature survey and expert interviews to explore how BBP and RD can facilitate the practice of identifying, classifying, prioritizing, remediating, and mitigating IoT vulnerabilities in an effective and cost-efficient manner. Besides deriving tangible guidelines for IoT stakeholders, our study also sheds light on a systematic integration path to combine BBP and RD with existing security practices (e.g., penetration test) to further boost overall IoT security.