Low-level Implementation Research Articles

Implementation of polygenic risk scores in secondary prevention of (breast) cancer

Abstract Utilizing polygenic risk scores to inform secondary prevention of breast cancer via population screening programs is one of the pioneering applications of genomics in public health applications. Over the last decade, polygenic risk scores for breast cancer have been developed and validated in various European populations (AUC ± 0.63, OR per Z unit ± 1.6, depending on the population). In some countries, the polygenic risk score has been integrated with decision modeling tools, such as CanRisk in the Netherlands, and low-level implementation into health and care systems has started. Implementation strategies vary, but generally consider initial implementation into high-risk populations, such as breast cancer families, in specialized clinical setting, such as academic medical centres, or specialized cancer clinics, referring relatives of patients to population screening based on comprehensive risk modelling, including polygenic risk scores. This route may be followed by other disease fields as well. The presentations outlines the current evidence and considerations on breast cancer genetic risk modelling, clinical pilot studies and the outline of step-wise implementation of such models to benefit secondary prevention programs. We start from the perspective of a single hospital, to the needed steps to upscale to national populations, both in diversity of these populations, as the throughput of genetic testing and counselling, until the harmonization and recalibration of various European healthcare models. We draw from examples of the Genotyping on all patients (GOALL), building cancer health platforms (CanHeal) and Genome of Europe (GoE) consortia.

Implementation Challenges of Building Information Modelling (BIM) in Small to Medium-Sized Enterprises (SMEs) Participating in Public Projects in Qatar

This research investigates the implementation challenges of Building Information Modelling (BIM) in Small to Medium-Sized Enterprises (SMEs) participating in public projects, with a focus on the construction industry in Qatar. Through a combination of literature review, surveys, and interviews, the study identifies key obstacles such as interoperability issues, financial constraints, organisational and cultural resistance, and the lack of comprehensive training programs. The literature highlights the fragmented nature of construction projects and the potential of BIM to improve efficiency through enhanced data handling and collaboration. However, SMEs often struggle with the high initial costs of software and hardware, as well as the need for extensive training to bridge the skill gap. Interviews with industry professionals reveal that while mandatory BIM integration in public projects drives adoption, it also exacerbates these challenges for SMEs, which may lack the necessary resources and expertise. Survey findings indicate that contractors frequently encounter issues with understanding and implementing BIM and meeting the government body standards, suggesting a need for better training and support. The study also emphasises the importance of strategic incremental adoption, where SMEs gradually integrate BIM into their processes, starting with low-level implementation in privately funded projects. The research concludes that while BIM offers significant benefits for the construction industry, its successful adoption requires targeted support for SMEs. This includes financial incentives, improved interoperability standards, comprehensive training programs, and strategic change management approaches. Addressing these challenges will enable SMEs to fully leverage BIM's potential, enhancing project efficiency and competitiveness in the construction sector.

Synchronous Programming with Refinement Types

Cyber-Physical Systems (CPS) consist of software interacting with the physical world, such as robots, vehicles, and industrial processes. CPS are frequently responsible for the safety of lives, property, or the environment, and so software correctness must be determined with a high degree of certainty. To that end, simply testing a CPS is insufficient, as its interactions with the physical world may be difficult to predict, and unsafe conditions may not be immediately obvious. Formal verification can provide stronger safety guarantees but relies on the accuracy of the verified system in representing the real system. Bringing together verification and implementation can be challenging, as languages that are typically used to implement CPS are not easy to formally verify, and languages that lend themselves well to verification often abstract away low-level implementation details. Translation between verification and implementation languages is possible, but requires additional assurances in the translation process and increases software complexity; having both in a single language is desirable. This paper presents a formalization of MARVeLus, a CPS language which combines verification and implementation. We develop a metatheory for its synchronous refinement type system and demonstrate verified synchronous programs executing on real systems.

LeanStore: A High-Performance Storage Engine for NVMe SSDs

Neither traditional disk-based database systems nor modern inmemory database systems are capable of fully exploiting modern servers with multiple NVMe SSDs. LeanStore is a high-performance OLTP storage engine specifically optimized for NVMe SSDs and multi-core CPUs. The paper gives an overview of the architecture of LeanStore and describes all major components, covering caching, page replacement, I/O management, indexing, data structure synchronization, multi-version concurrency control, logging, checkpoints, and recovery. We also discuss some of the low-level implementation techniques necessary for achieving high performance on modern hardware.

A verified low-level implementation and visualization of the adaptive exterior light and speed control system

In this article, we present an approach to the ABZ 2020 case study that differs from those usually presented at ABZ: Rather than using a (correct-by-construction) approach following a formal method, we use C for a low-level implementation instead. We strictly adhere to test-driven development for validation, and only afterwards apply model checking using CBMC for verification. While the approach has several benefits compared to the more rigorous approaches, it also provides less mathematical clarity and overall less thorough verification. In consequence, our realization of the ABZ case study serves as a baseline reference for comparison, allowing to assess the benefit provided by the various formal modeling languages, methods and tools.

Developing CBTC system safety requirement hierarchy through STPA methodology

Safety-critical systems, such as the railway signal system, are subject to potentially high costs from failures, including loss of life and property damage. The use of new technology, including communication-based train control (CBTC) systems with software and computers, has changed the types of accidents that occur. Software-related issues and dysfunctional interactions between system components controlled by the software are increasingly the cause of incidents. Developing a “safe” safety-critical system requires accurate and complete safety requirements, which are the foundation of system development. Traditional hazard analysis techniques are insufficient for identifying the causes of accidents in modern railway signaling systems. Systems-Theoretic Process Analysis (STPA) is a powerful new hazard analysis method designed to address these limitations. Building upon this foundation, a hierarchical approach to safety requirement development has been further developed. This approach combines STPA analysis with a hierarchical modeling approach to establish traceability links from safety requirements to specific architectures, refine and allocate system-level safety requirements to relevant subsystems, and abstract safety requirements at higher hierarchical levels to enable easy changes to lower-level implementations. This paper employs the aforementioned methodology within the context of the CBTC system, thereby enhancing risk management and hazard analysis, enabling early insights, and facilitating the generation of safety requirements of CBTC System.

SMaLL: Software for Rapidly Instantiating Machine Learning Libraries

Interest in deploying deep neural network (DNN) inference on edge devices has resulted in an explosion of the number and types of hardware platforms that machine learning (ML) libraries must support. High-level programming interfaces, such as TensorFlow, can be readily ported across different devices; however, maintaining performance when porting the low-level implementation is more nuanced. High-performance inference implementations require an effective mapping of the high-level interface to the target hardware platform. Commonly, this mapping may use optimizing compilers to generate code at compile time or high-performance vendor libraries that have been specialized to the target platform. Both approaches rely on expert knowledge across levels to produce an efficient mapping. This makes supporting new architectures difficult and time-consuming. In this work, we present a DNN library framework, SMaLL, that is easily extensible to new architectures. The framework uses a unified loop structure and shared, cache-friendly data format across all intermediate layers, eliminating the time and memory overheads incurred by data transformation between layers. Each layer is implemented by specifying its dimensions and a kernel , the key computing operation of that layer. The unified loop structure and kernel abstraction allows the reuse of code across layers and computing platforms. New architectures only require a few hundred lines in the kernel to be redesigned. To show the benefits of our approach, we have developed software that supports a range of layer types and computing platforms; this software is easily extensible for rapidly instantiating high-performance DNN libraries. An evaluation of the portability of our framework is shown by instantiating end-to-end networks from the MLPerf:tiny benchmark suite on five ARM platforms and one x86 platform (an AMD Zen 2). We also show that the end-to-end performance is comparable to or better than ML frameworks such as TensorFlow, TVM, and LibTorch.

Implementation and Synthesis of Math Library Functions

Achieving speed and accuracy for math library functions like exp, sin, and log is difficult. This is because low-level implementation languages like C do not help math library developers catch mathematical errors, build implementations incrementally, or separate high-level and low-level decision making. This ultimately puts development of such functions out of reach for all but the most experienced experts. To address this, we introduce MegaLibm, a domain-specific language for implementing, testing, and tuning math library implementations. MegaLibm is safe, modular, and tunable. Implementations in MegaLibm can automatically detect mathematical mistakes like sign flips via semantic wellformedness checks, and components like range reductions can be implemented in a modular, composable way, simplifying implementations. Once the high-level algorithm is done, tuning parameters like working precisions and evaluation schemes can be adjusted through orthogonal tuning parameters to achieve the desired speed and accuracy. MegaLibm also enables math library developers to work interactively, compiling, testing, and tuning their implementations and invoking tools like Sollya and type-directed synthesis to complete components and synthesize entire implementations. MegaLibm can express 8 state-of-the-art math library implementations with comparable speed and accuracy to the original C code, and can synthesize 5 variations and 3 from-scratch implementations with minimal guidance.

Toward Efficient Convolutional Neural Networks With Structured Ternary Patterns.

High-efficiency deep learning (DL) models are necessary not only to facilitate their use in devices with limited resources but also to improve resources required for training. Convolutional neural networks (ConvNets) typically exert severe demands on local device resources and this conventionally limits their adoption within mobile and embedded platforms. This brief presents work toward utilizing static convolutional filters generated from the space of local binary patterns (LBPs) and Haar features to design efficient ConvNet architectures. These are referred to as Structured Ternary Patterns (STePs) and can be generated during network initialization in a systematic way instead of having learnable weight parameters thus reducing the total weight updates. The ternary values require significantly less storage and with the appropriate low-level implementation, can also lead to inference improvements. The proposed approach is validated using four image classification datasets, demonstrating that common network backbones can be made more efficient and provide competitive results. It is also demonstrated that it is possible to generate completely custom STeP-based networks that provide good trade-offs for on-device applications such as unmanned aerial vehicle (UAV)-based aerial vehicle detection. The experimental results show that the proposed method maintains high detection accuracy while reducing the trainable parameters by 40%-80%. This work motivates further research toward good priors for nonlearnable weights that can make DL architectures more efficient without having to alter the network during or after training.

“Sometimes it can be like an icebreaker”: A mixed method evaluation of the implementation of the Refugee Health Screener-13 (RHS-13)

BackgroundForced migrants are at risk of developing mental illness, yet challenges remain with underutilization of mental healthcare among this population. This study examined the implementation of the Refugee Health Screener-13 (RHS-13) in the health assessment for forced migrants in eight primary health care centres in Stockholm Region, Sweden. MethodsA mixed-methods convergent parallel design was used, combining nurses self-reported quantitative data on the levels and reasons for RHS-13 use in the health assessment with qualitative interview data on the barriers and facilitators for RHS-13 use. The Consolidated Framework for Implementation Research (CFIR) was used as a coding framework for the qualitative analysis. ResultsLevels of RHS-13 use varied between primary health care centres, resulting in two groups: three centres with high-level (65–92%) and five centres with low-level (0–36%) implementation. Factors related to the tool itself, as well as the inner and outer context, influenced the use of RHS-13. Language barriers, insufficient time, and lack of trust in the validity and utility of RHS-13 were the main barriers, while its availability in many languages and that it was perceived as an important complement to the health assessment were the main facilitators. ConclusionRHS-13 contributes to the standardization of assessing mental health in the health assessment. Identifying context-based implementation strategies and addressing language and time issues as well as nurses trust in the tool's utility are recommended to enhance the use of RHS-13.

VisEval: A Benchmark for Data Visualization in the Era of Large Language Models.

Translating natural language to visualization (NL2VIS) has shown great promise for visual data analysis, but it remains a challenging task that requires multiple low-level implementations, such as natural language processing and visualization design. Recent advancements in pre-trained large language models (LLMs) are opening new avenues for generating visualizations from natural language. However, the lack of a comprehensive and reliable benchmark hinders our understanding of LLMs' capabilities in visualization generation. In this paper, we address this gap by proposing a new NL2VIS benchmark called VisEval. Firstly, we introduce a high-quality and large-scale dataset. This dataset includes 2,524 representative queries covering 146 databases, paired with accurately labeled ground truths. Secondly, we advocate for a comprehensive automated evaluation methodology covering multiple dimensions, including validity, legality, and readability. By systematically scanning for potential issues with a number of heterogeneous checkers, VisEval provides reliable and trustworthy evaluation outcomes. We run VisEval on a series of state-of-the-art LLMs. Our evaluation reveals prevalent challenges and delivers essential insights for future advancements.

A review of mathematical and computational aspects of CSIDH algorithms

CSIDH is a post-quantum key-exchange scheme based on the action of ideal class groups on supersingular elliptic curves over prime fields. Its short keys and ciphertexts, together with its flexibility as a building block to construct complex cryptographic primitives, has motivated significant research on the efficiency of CSIDH and its resistance against side-channel attacks. In this work, some cutting-edge results from recent contributions are reviewed in a unified treatment, focusing on the mathematical ideas lying behind them rather than on cryptographic and low-level implementation techniques. In particular, we first describe ways to speed up the class-group-action evaluation, which range from the use of different models of elliptic curves to working with different ideal class groups. We then survey some constant-time variants of CSIDH, that make the time and memory consumption during the computation of a public/shared key independent of the secret key. Finally, we examine the computation of the ideal class action when the structure of the ideal class group is known, which is the case for a specific set of CSIDH parameters.

Counterexample Driven Quantifier Instantiations with Applications to Distributed Protocols

Formally verifying infinite-state systems can be a daunting task, especially when it comes to reasoning about quantifiers. In particular, quantifier alternations in conjunction with function symbols can create function cycles that result in infinitely many ground terms, making it difficult for solvers to instantiate quantifiers and causing them to diverge. This can leave users with no useful information on how to proceed. To address this issue, we propose an interactive verification methodology that uses a relational abstraction technique to mitigate solver divergence in the presence of quantifiers. This technique abstracts functions in the verification conditions (VCs) as one-to-one relations, which avoids the creation of function cycles and the resulting proliferation of ground terms. Relational abstraction is sound and guarantees correctness if the solver cannot find counter-models. However, it may also lead to false counterexamples, which can be addressed by refining the abstraction and requiring the existence of corresponding elements. In the domain of distributed protocols, we can refine the abstraction by diagnosing counterexamples and manually instantiating elements in the range of the original function. If the verification conditions are correct, there always exist finitely many refinement steps that eliminate all spurious counter-models, making the approach complete. We applied this approach in Ivy to verify the safety properties of consensus protocols and found that: (1) most verification goals can be automatically verified using relational abstraction, while SMT solvers often diverge when given the original VC, (2) only a few manual instantiations were needed, and the counterexamples provided valuable guidance for the user compared to timeouts produced by the traditional approach, and (3) the technique can be used to derive efficient low-level implementations of tricky algorithms.

Code semantic enrichment for deep code search

Code search aims to retrieve code snippets from a large-scale codebase, where the semantics of the searched code match developers’ query intent. Code is a low-level implementation of programming intents, but query is always expressed as clear and high-level semantics, which makes it difficult for DL-based approaches to learn the semantic relationship between them. Through a large-scale empirical analysis on more than 2.2 million pairs of Java code and description, we found that the semantics of code and query can be aligned by enriching code with the descriptions of other code in terms of similar implementation. Based on the finding, we propose a code semantic enrichment approach for deep code search, named SemEnr. Specifically, we first enrich semantics for all code snippets in the training and testing data. We estimated the syntactic similarity of each code snippet from the training data and retrieved the most similar one for each. Thereafter, the semantics of one code snippet is represented by its code tokens and the description of the retrieved most similar code. During the model training, we used the attention mechanism to embed pairs of enriched code and query into the shared high-dimensional vector space. To enhance the quality of our learned representations, we integrated a multi-perspective co-attention mechanism, employing Convolutional Neural Networks (CNNs) to capture local correlations between code and query. Finally, we evaluated the effectiveness of our approach by performing experiments on two extensively used Java datasets. Our experimental results reveal that SemEnr achieves an MRR of 0.698 and 0.631, outperforming the best baseline CAT (a state-of-the-art DL-based model) by 19.93% and 18.83%, respectively. In addition, we conducted a user study involving 50 real-world queries to assess SemEnr’s performance, and the findings suggest that SemEnr outperformed baseline models by returning more relevant code snippets.

Entering the Metaverse from the JVM: The State of the Art, Challenges, and Research Areas of JVM-Based Web 3.0 Tools and Libraries

Web 3.0 is the basis on which the proposed metaverse, a seamless virtual world enabled by computers and interconnected devices, hopes to interact with its users, but beyond the high-level project overview of what Web 3.0 applications try to achieve, the implementation is still down to low-level coding details. This article aims to analyze the low-level implementations of key components of Web 3.0 using a variety of frameworks and tools as well as several JVM-based languages. This paper breaks down the low-level implementation of smart contracts and semantic web principles using three frameworks, Corda and Ethereum for smart contracts and Jeda for semantic web, using both Scala and Java as implementing languages all while highlighting differences and similarities between the frameworks used.

Simulation-supported development for cooperative Multi-UAV Systems with the Mysterio framework

Over the years, UAVs (also known as drones) have been growing in studies and applications to solve diverse problems. Due to the complexity of these problems, dealing with just one UAV may not be enough, but using several UAVs together to work cooperatively increases its capacities, thus boosting innovative solutions. However, developing cooperative Multi-UAV systems is not trivial, and reuse support is usually limited to low-level implementation. This work presents a framework for Multi-UAVs, called Mysterio, which provides an underlying software architecture with essential Multi-UAV components, enabling the reuse of design and code so that engineers can instantiate it to carry out specific missions by making UAVs work in cooperation. We also present four instances of the framework to evaluate Mysterio’s effectiveness in different scenarios. Finally, we discuss the framework’s potential to provide and support design and code reuse to develop Cooperative Multi-UAVs systems for different application scenarios. The results showed the potential to develop multi-UAV systems using the proposed framework. Additionally, we extend our previous work bringing conceptual evolution and advances in the architecture and the framework. Finally, this evolution extends the framework API to support computer simulations of UAV systems based on the OMNeT++ simulator. This API is suitable for Single-UAV and Multi-UAV systems and has already been adapted to communicate with base stations implemented through the Mysterio Framework.

Design Space Exploration for Partitioning Dataflow Program on CPU-GPU Heterogeneous System

Dataflow programming is a methodology that enables the development of high-level, parametric programs that are independent of the underlying platform. This approach is particularly useful for heterogeneous platforms, as it eliminates the need to rewrite application software for each configuration. Instead, it only requires new low-level implementation code, which is typically automatically generated through code generation tools. The performance of programs running on heterogeneous parallel platforms is highly dependent on the partitioning and mapping of computation to different processing units. This is determined by parameters that govern the partitioning, mapping, scheduling, and allocation of data exchanges among the processing elements of the platform. Determining the appropriate parameters for a specific application and set of architectures is a complex task and is an active area of research. This paper presents a novel methodology for partitioning and mapping dataflow programs onto heterogeneous systems composed of both CPUs and GPUs. The objective is to identify the program configuration that provides the most efficient way to process a typical dataflow program by exploring its design space. This is an NP-complete problem that we have addressed by utilizing a design space exploration approach that leverages a Tabu search meta-heuristic optimization algorithm driven by analysis of the execution trace graph of the program. The heuristic algorithm effectively identifies a solution that maps actors to processing units while improving overall performance. The parameters of the heuristic algorithm, such as the time limit and the proportion of neighboring solutions explored during each iteration, can be fine-tuned for optimal results. Additionally, the proposed approach allows for the exploration of solutions that do not utilize all hardware resources if it results in better performance. The effectiveness of the proposed approach is demonstrated through experimental results on dataflow programs.

When is universal SEL effective under authentic conditions? Using LPA to examine program implementation in elementary classrooms

As universal social-emotional learning (SEL) programs have become more common in K-12 schools, implementation practices have been found to affect program quality. However, research examining how multiple facets of program implementation interrelate and impact student outcomes, especially under routine conditions in schools, is still limited. As such, we used latent profile analysis (LPA) to examine implementation of a brief universal SEL program (Social Skills Improvement System SEL Classwide Intervention Program) in primary classrooms. Three latent profiles of implementation were identified based on dosage, adherence, quality of delivery, student engagement and teachers’ impression of lessons. Although results suggested that classrooms with moderate- and high-level implementation practices generally showed higher gains in student outcomes than those with low-level implementation, these differences did not reach statistical significance except for academic motivation. Implications for school-based universal SEL program planning, implementation, and evaluation are discussed.

Using a single-channel reference with the MBSTOI binaural intelligibility metric

In order to assess the intelligibility of a target signal in a noisy environment, intrusive speech intelligibility metrics are typically used. They require a clean reference signal to be available which can be difficult to obtain especially for binaural metrics like the modified binaural short time objective intelligibility metric (MBSTOI). We here present a hybrid version of MBSTOI that incorporates a deep learning stage that allows the metric to be computed with only a single-channel clean reference signal. The models presented are trained on simulated data containing target speech, localised noise, diffuse noise, and reverberation. The hybrid output metrics are then compared directly to MBSTOI to assess performances. Results show the performance of our single channel reference vs MBSTOI. The outcome of this work offers a fast and flexible way to generate audio data for machine learning (ML) and highlights the potential for low level implementation of ML into existing tools.

Cooperative control and stability analysis for virtual coupling of rail vehicles

Cooperative control for rail vehicles is a new idea which builds on existing models and methods developed in the context of flight formation and autonomous ground vehicles. The underlying idea is to assimilate virtually coupled rail vehicles to arrays of mass–spring–dashpot systems. The contribution of this paper is three-fold. First, a novel scheme is proposed which involves a supervisory control for the higher-level planning and a local motion coordination for the lower-level implementation. For the supervisory control, a review of the existing scheduling models and methods based on alternate graphs and a mixed-integer (non)linear program is conducted. For the local coordination, a nonlinear and uncertain model is developed and a constructive method is provided to design a feedback control that stabilizes the vehicle around a desired equilibrium point, in terms of position and velocity. Second, the control design method is extended to the case of multiple virtually coupled vehicles. It is proven that the transient dynamics follows a typical synchronization dynamics. It is also proven that under such control, the whole system of rail vehicles converges to a pre-defined equilibrium point, characterized by a specific velocity and relative distance between vehicles. Conditions for the stability of such equilibrium points are investigated. Third, under the hypothesis of homogeneity, bounds on the damping coefficient for the synchronization equilibrium (in terms of velocity and relative distance between vehicles) to be overdamped or underdamped are provided.