Search for: All records

The rapid integration of AI components into software systems introduces new complexities for requirements engineering (RE). Despite the widespread adoption of AI, few public datasets of requirements and their associated trace links exist for these systems. This lack of available data hinders progress in RE4AI, limiting both the evaluation of existing techniques and the development of new approaches that address the unique challenges of AI. To help bridge this gap, we present a dataset from a real-world Generative AI application consisting of source code written in multiple languages and a diverse hierarchy of software artifacts, including features, functional and non-functional requirements, and design constraints. The dataset also provides manually constructed trace links between these natural language artifacts, with end-to-end traceability to code for a subset of AI-related features. 

Cyber-Physical Systems (CPS) increasingly leverage Reinforcement Learning (RL) to adapt dynamically to changing environments and optimize performance over time. While RL enhances efficiency and safety by enabling autonomous adjustments to unexpected conditions and hazard avoidance, it also introduces significant risks, as learned behaviors may lead to unpredictable or unsafe actions in real-world deployment. Therefore, integrating risk management into RL system design is essential. In this paper, we propose the \framework, a question-driven approach that translates high-level safety guidelines into RL-specific considerations. This framework helps RL practitioners address key risks early in development, informing new or existing system requirements while ensuring traceability to risk management objectives. To evaluate its effectiveness, we conducted a study across two use cases, engaging six RL experts in developing system requirements with and without the framework. Our findings suggest that the framework promotes critical thinking and helps practitioners identify additional risk factors, ultimately supporting safer RL deployment. 

Small uncrewed aerial systems, sUAS, provide an invaluable resource for performing a variety of surveillance, search, and delivery tasks in remote or hostile terrains which may not be accessible by other means. Due to the critical role sUAS play in these situations, it is vital that they are well configured in order to ensure a safe and stable flight. However, it is not uncommon for mistakes to occur in configuration and calibration, leading to failures or incomplete missions. To address this problem, we propose a set of self-adaptive mechanisms and implement them into a self-adaptive framework,CICADA, for Controller Instability-preventing Configuration Aware Drone Adaptation.CICADAdynamically detects unstable drone behavior during flight and adapts to mitigate this threat. We have built a prototype ofCICADAusing a popular open source sUAS flight control software and experimented with a large number of different configurations in simulation. We then performed a case study with physical drones to determine if our framework will work in practice. Experimental results show thatCICADA’sadaptations reduce controller instability and enable the sUAS to recover from up to 33.8% of poor configurations. In cases where we cannot complete the intended mission, invoking alternative adaptations may still help by allowing the vehicle to loiter or land safely in place, avoiding potentially catastrophic crashes. These safety-focused adaptations can mitigate unsafe behavior in 52.9% to 64.7% of dangerous configurations. We further show that rule-based approaches can be leveraged to automatically select an appropriate adaptation strategy based on the severity of instability encountered, with up to a 14.2% improvement over direct adaptation. Finally, we introduce a variation of our primary adaptation strategy designed to allow more cautious adaptation with limited configuration information, which gets within 6.7% of our primary adaptation strategy despite not requiring an optimal knowledge base. 

Software documentation supports a broad set of software maintenance tasks; however, creating and maintaining high-quality, multi-level software documentation can be incredibly time-consuming and therefore many code bases suffer from a lack of adequate documentation. We address this problem through presenting HGEN, a fully automated pipeline that leverages LLMs to transform source code through a series of six stages into a well-organized hierarchy of formatted documents. We evaluate HGEN both quantitatively and qualitatively. First, we use it to generate documentation for three diverse projects, and engage key developers in comparing the quality of the generated documentation against their own previously produced manually-crafted documentation. We then pilot HGEN in nine different industrial projects using diverse datasets provided by each project. We collect feedback from project stakeholders, and analyze it using an inductive approach to identify recurring themes. Results show that HGEN produces artifact hierarchies similar in quality to manually constructed documentation, with much higher coverage of the core concepts than the baseline approach. Stakeholder feedback highlights HGEN's commercial impact potential as a tool for accelerating code comprehension and maintenance tasks. 

Software engineering practices such as constructing requirements and establishing traceability help ensure systems are safe, reliable, and maintainable. However, they can be resource-intensive and are frequently underutilized. To alleviate the burden of these essential processes, we developed the Requirements Organization and Optimization Tool (ROOT). ROOT centralizes project information and offers project visualizations and AI-based tools designed to streamline engineering processes. With ROOT's assistance, engineers benefit from improved oversight and early error detection, leading to the successful development of software systems. 

Software documentation supports a broad set of software maintenance tasks; however, creating and maintaining high-quality, multi-level software documentation can be incredibly time-consuming and therefore many code bases suffer from a lack of adequate documentation. We address this problem through presenting HGEN, a fully automated pipeline that leverages LLMs to transform source code through a series of six stages into a well-organized hierarchy of formatted documents. We evaluate HGEN both quantitatively and qualitatively. First, we use it to generate documentation for three diverse projects, and engage key developers in comparing the quality of the generated documentation against their own previously produced manually-crafted documentation. We then pilot HGEN in nine different industrial projects using diverse datasets provided by each project. We collect feedback from project stakeholders, and analyze it using an inductive approach to identify recurring themes. Results show that HGEN produces artifact hierarchies similar in quality to manually constructed documentation, with much higher coverage of the core concepts than the baseline approach. Stakeholder feedback highlights HGEN's commercial impact potential as a tool for accelerating code comprehension and maintenance tasks. Results and associated supplemental materials can be found at https://zenodo.org/records/11403244. 

Software engineering practices such as constructing requirements and establishing traceability help ensure systems are safe, reliable, and maintainable. However, they can be resource-intensive and are frequently underutilized. To alleviate the burden of these essential processes, we developed the Requirements Organization and Optimization Tool (ROOT). ROOT centralizes project information and offers project visualizations and AI-based tools designed to streamline engineering processes. With ROOT's assistance, engineers benefit from improved oversight and early error detection, leading to the successful development of software systems. A link to a screen cast can be found at: https://youtu.be/3rtMYRnsu24 

Data associated with Huang et al., "High Resolution ALMA Observations of Richly Structured Protoplanetary Disks in σ Orionis," accepted by ApJ.  The raw data can be obtained from the ALMA archive under program IDs 2016.1.00447.S (PI: J. Williams) and 2022.1.00728.S (PI: J. Huang).  Contents images.tar: FITS files of the ALMA continuum images of the eight disks reductionscripts.tar: CASA reduction scripts visibilities.tar: Continuum measurement sets for the eight disks (note that the weights in these measurement sets are not rescaled) 

Abstract The Atacama Large Millimeter/submillimeter Array (ALMA) has detected substructures in numerous protoplanetary disks at radii from a few to over 100 au. These substructures are commonly thought to be associated with planet formation, either by serving as sites fostering planetesimal formation or by arising as a consequence of planet–disk interactions. Our current understanding of substructures, though, is primarily based on observations of nearby star-forming regions with mild UV environments, whereas stars are typically born in much harsher UV environments, which may inhibit planet formation in the outer disk through external photoevaporation. We present high-resolution (∼8 au) ALMA 1.3 mm continuum images of eight disks inσOrionis, a cluster irradiated by an O9.5 star. Gaps and rings are resolved in the images of five disks. The most striking of these is SO 1274, which features five gaps that appear to be arranged nearly in a resonant chain. In addition, we infer the presence of gap or shoulder-like structures in the other three disks through visibility modeling. These observations indicate that substructures robustly form and survive at semimajor axes of several tens of au or less in disks exposed to intermediate levels of external UV radiation as well as in compact disks. However, our observations also suggest that disks inσOrionis are mostly small, and thus millimeter continuum gaps beyond a disk radius of 50 au are rare in this region, possibly due to either external photoevaporation or age effects. 

Recent innovations in virtual and mixed-reality (VR/MR) technologies have enabled innovative hands-on training applications in high-risk/high-value fields such as medicine, flight, and worker-safety. Here, we present a detailed description of a novel VR/MR tactile user interactions/interface (TUI) hardware and software development framework that enables the rapid and cost-effective no-code development, optimization, and distribution of fully authentic hands-on VR/MR laboratory training experiences in the physical and life sciences. We applied our framework to the development and optimization of an introductory pipette calibration activity that is often carried out in real chemistry and biochemistry labs. Our approach provides users with nuanced real-time feedback on both their psychomotor skills during data acquisition and their attention to detail when conducting data analysis procedures. The cost-effectiveness of our approach relative to traditional face-to-face science labs improves access to quality hands-on science lab experiences. Importantly, the no-code nature of this Hands-On Virtual-Reality (HOVR) Lab platform enables faculties to iteratively optimize VR/MR experiences to meet their student’s targeted needs without costly software development cycles. Our platform also accommodates TUIs using either standard virtual-reality controllers (VR TUI mode) or fully functional hand-held physical lab tools (MR TUI mode). In the latter case, physical lab tools are strategically retrofitted with optical tracking markers to enable tactile, experimental, and analytical authenticity scientific experimentation. Preliminary user study data highlights the strengths and weaknesses of our generalized approach regarding student affective and cognitive student learning outcomes.