Preventing unexpected long-latency spikes is crucial for latency-sensitive hardware systems like Solid-State Drives (SSDs). Conventional test case (TC) generation methods often lack reproducibility and rely on proprietary internal firmware knowledge, limiting their applicability. To address this, we propose SNIFFER, a vendor-agnostic, black-box framework that utilizes Reinforcement Learning (RL) for the automated generation of latency anomaly-inducing TCs. SNIFFER interacts directly with real SSD hardware, using only externally observable metrics from standardized tools like Flexible I/O and Open Compute Project. Our framework formulates the problem as a sequential decision-making process, enabling an RL agent to learn complex I/O patterns that induce stress. We demonstrate that SNIFFER consistently generates effective TCs, inducing up to 74.7% higher maximum latency in up to 85% fewer steps compared to a random baseline. More importantly, we demonstrate its superiority over alternative black-box optimization methods, such as Genetic Algorithms, validating our approach for non-stationary hardware environments. SNIFFER ’s ability to reproducibly generate diverse and stressful TCs makes it a powerful tool for automated industrial validation pipelines.

Well-designed test cases (TCs) are crucial for en-suring the quality of Solid-State Drive (SSD) products. Indeed, validating SSD firmware code by the TCs is indispensable to check if there are no defects during the SSD development process. Accordingly, it is necessary to create short TCs covering firmware code as much as possible for efficient and precise validation. While various methods are available for generating TCs, existing automated approaches overlook backward compatibility, a key property in the SSD development process. To utilize the property, we introduce a novel deep-learning approach called GRAFT, which combines graph representation learning and reinforcement learning (RL) for automated TC generation in SSDs by leveraging pre-collected data. G RAFT trains a graph neural network to extract the underlying structure of the SSD firmware code from a given SSD simulator. The resulting graph embeddings serve as observations in the RL process. To address the challenge of over- estimation in an external domain in the RL process, conservative Q-Iearning, an offline RL technique, is employed using the pre- collected data. Despite the limitation of not being able to interact with the SSD simulator for training, we demonstrate that GRAFT successfully trains RL agents that generate TCs. The TCs are not only significantly more efficient with 3.5x shorter than randomly generated TCs by a black-box fuzzer but also exhibit comparable coverage and efficiency to those created by human experts with domain knowledge, which fully took three days. Moreover, the TCs achieves maximum coverage more reliably than any other methods in the experiments.

Urban traffic congestion has become an important issue not only affecting our daily lives, but also limiting economic development. The primary cause of urban traffic congestion is that the number of vehicles is higher than the permissible limit of the road. Previous studies have focused on dispersing traffic volume by detecting urban traffic congestion zones and predicting future trends. However, to solve the fundamental problem, it is necessary to discover the cause of traffic congestion. Nevertheless, it is difficult to find a research which presents an approach to identify the causes of traffic congestion. In this paper, we propose a technique to analyze the cause of traffic congestion based on the traffic flow theory. We extract vehicle flows from traffic data, such as GPS trajectory and Vehicle Detector data. We detect vehicle flow changes utilizing the entropy from the information theory. Then, we build cumulative vehicle count curves (N-curve) that can quantify the flow of the vehicles in the traffic congestion area. The N-curves are classified into four different traffic congestion patterns by a convolutional neural network. Analyzing the causes and influence of traffic congestion is difficult and requires considerable experience and knowledge. Therefore, we present a visual analytics system that can efficiently perform a series of processes to analyze the cause and influence of traffic congestion. Through case studies, we have evaluated that our system can classify the causes of traffic congestion and can be used efficiently in road planning.

It is not easy to predict incomplete panel data whose overall trend is not complete. For example, physical growth data for age 7-18 has been collected every a few months for up to seven years; therefore, only short growth pattern pieces exist in the data. When using previous prediction techniques, it is challenging to create a growth prediction model that reflects individual growth patterns. Also, uncertainties in predicted results emerge since the overall trend of the data is unknown. In this work, we present a predictive analysis and modeling to forecast incomplete data over time. We extend the Bayesian network model, which is relative data-driven approaches that explore similar data and weave them to create approximate inference margins. Besides, we propose a visual analytics system that enables us to design various predictive models that reflect individual growth pattern. Our visual analytics system assists us to discover new growth patterns in the process of analyzing the accuracy of previously designed predictive models. Moreover, the system allows us to optimize predictive models to fit unusual growth patterns.