Existing Software Frameworks Research Articles

HW/SW Codesign for Approximation-Aware Binary Neural Networks

Binary Neural Networks (BNNs) are rapidly gaining remarkable attention due to their superiority in shrinking the model size, which outstandingly mitigates the fundamental “memory wall” bottleneck that is attributed to the existing von-Neumann architectures. This work investigates how principles from approximate computing can be effectively employed to further optimize BNNs. It demonstrates that HW/SW codesign, in which BNNs are either proactively trained in the presence of approximation-induced errors (i.e. design-time optimization) and/or augmented with an appropriate error-mitigation scheme (i.e., run-time optimization), is a key to realize energy-efficient yet robust BNNs. We unveil, for the first time, that although the underlying HW of BNNs can be implemented using simple <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">XNOR</monospace> gates, the complexity of the required “Popcount” circuit super-linearly grows with the filter kernel size. This largely impacts the area footprint, inference time, energy, and hence it severely constricts the prospective efficiency gains from BNNs. To overcome this challenge, we replace the accurate full adders constructing the Popcount with Majority gates that approximately perform the required additions. Then, our carefully-crafted error-mitigation scheme along with activations tuning considerably minimizes the induced errors. Afterward, abstracted error probabilities are derived and employed during BNN training to obtain approximation-aware BNNs, that are inherently robust against the underlying HW approximation. Differently from the typical approaches, the proposed HW/SW codesign methodology has the merit of allowing a training of the approximate BNN without the need to modify the existing software frameworks (i.e., PyTorch). This is of great importance since existing tools rely on efficient built-in functions that can be difficult and/or inefficient to be modified. An FPGA-based SoC realizing both accurate and approximation-aware BNNs is developed for validating our proposed methodology. With merely a 4.7% loss in the inference accuracy, our HW/SW codesign leads to 64% and 80.2% savings in the area and energy, respectively, at the parity of the latency. Our results are obtained using commercial EDA tool flows employing a commercial 28nm FDSOI technology node.

Towards a Novel Software Framework for the Intuitive Generation of Process Flows for Multiple Robotic Systems

To make the process-specific programming of robotic and technical systems more efficient, we introduce the concept of a novel software framework to enable intuitive configuration of cooperative process flows for multiple systems. Our concept allows both the easy programming of systems and the generation of process flows with multiple systems to enable cooperation between them. Following the idea of Operator 4.0 our framework thus enables non-expert workers of industrial companies to effectively control the process flows of a production or logistical process with minimal training effort and costs.In this paper, we first analyze existing software frameworks that facilitate graphical programming and differentiate our approach from them. We explain in detail how our approach conceptually and technically works regarding the framework´s technical architecture, system requirements, interface design, user assistance methods based on digital twin and data analysis, as well as interaction patterns.We validate the functionality and generality of our software framework by integrating two systems of different manufacturers – a collaborative robot and a robotic gripper – into our framework and discuss the technical capabilities of our prototype. In addition, the interaction of unskilled workers and our framework will be compared to an industrial robot teach pendant in a case study in which the participants had to create a pick-and-place process flow.

GraphAttack

Graph structures are a natural representation of important and pervasive data. While graph applications have significant parallelism, their characteristic pointer indirect loads to neighbor data hinder scalability to large datasets on multicore systems. A scalable and efficient system must tolerate latency while leveraging data parallelism across millions of vertices. Modern Out-of-Order (OoO) cores inherently tolerate a fraction of long latencies, but become clogged when running severely memory-bound applications. Combined with large power/area footprints, this limits their parallel scaling potential and, consequently, the gains that existing software frameworks can achieve. Conversely, accelerator and memory hierarchy designs provide performant hardware specializations, but cannot support diverse application demands. To address these shortcomings, we present GraphAttack, a hardware-software data supply approach that accelerates graph applications on in-order multicore architectures. GraphAttack proposes compiler passes to (1) identify idiomatic long-latency loads and (2) slice programs along these loads into data Producer/ Consumer threads to map onto pairs of parallel cores. Each pair shares a communication queue; the Producer asynchronously issues long-latency loads, whose results are buffered in the queue and used by the Consumer. This scheme drastically increases memory-level parallelism (MLP) to mitigate latency bottlenecks. In equal-area comparisons, GraphAttack outperforms OoO cores, do-all parallelism, prefetching, and prior decoupling approaches, achieving a 2.87× speedup and 8.61× gain in energy efficiency across a range of graph applications. These improvements scale; GraphAttack achieves a 3× speedup over 64 parallel cores. Lastly, it has pragmatic design principles; it enhances in-order architectures that are gaining increasing open-source support.

A Highly Modular Software Framework for Reducing Software Development Time of Nanosatellites

The standardization of the physical aspects of nanosatellites (also known as CubeSats) and their wide adoption in academia and industry has made the mass production and availability of off-the-shelf components possible. While this has led to a significant reduction in satellite development time, the fact remains that a considerable amount of mission development time and effort continues to be spent on flight software development. The CubeSat’s agile development environment makes it challenging to utilize the advantages of existing software frameworks. Such an adoption is not straightforward due to the added complexity characterized by a steep learning curve. A well-designed flight software architecture mitigates possible sources of failure and increases mission success rate while maintaining moderate complexity. This paper presents a novel approach to a flight software framework developed specifically for nanosatellites. The software framework is characterized by simplicity, reliability, modularity, portability, and real-time capability. The main features of the proposed framework include providing a standardized and explicit skeleton for each module to simplify their construction, offering standardized interfaces for all modules to simplify communication, and providing a collection of ready-to-use common services open for further enhancement by CubeSat software developers. The framework efficiency was demonstrated through a software developed for the MeznSat mission that was successfully launched into Low Earth Orbit in September 2020. The proposed software framework proved to simplify software development for the application developer while significantly enhancing software modularity.

DESDEO: The Modular and Open Source Framework for Interactive Multiobjective Optimization

Interactive multiobjective optimization methods incorporate preferences from a human decision maker in the optimization process iteratively. This allows the decision maker to focus on a subset of solutions, learn about the underlying trade-offs among the conflicting objective functions in the problem and adjust preferences during the solution process. Incorporating preference information allows computing only solutions that are interesting to the decision maker, decreasing computation time significantly. Thus, interactive methods have many strengths making them viable for various applications. However, there is a lack of existing software frameworks to apply and experiment with interactive methods. We fill a gap in the optimization software available and introduce DESDEO, a modular and open source Python framework for interactive multiobjective optimization. DESDEO’s modular structure enables implementing new interactive methods and reusing previously implemented ones and their functionalities. Both scalarization-based and evolutionary methods are supported, and DESDEO allows hybridizing interactive methods of both types in novel ways and enables even switching the method during the solution process. Moreover, DESDEO also supports defining multiobjective optimization problems of different kinds, such as data-driven or simulation-based problems. We discuss DESDEO’s modular structure in detail and demonstrate its capabilities in four carefully chosen use cases aimed at helping readers unfamiliar with DESDEO get started using it. We also give an example on how DESDEO can be extended with a graphical user interface. Overall, DESDEO offers a much-needed toolbox for researchers and practitioners to efficiently develop and apply interactive methods in new ways – both in academia and industry.

Simulation of natural disasters and managing rescue operations via geospatial crowdsourcing services in tensor space

Natural disasters have always been difficult to manage because of their spatio-temporal scale. Although the use of GIS and RS has made this process somewhat simpler, there are still serious challenges in managing and smart allocating human resources, increasing operational power and cost savings, continuous monitoring, and ensuring the effectiveness of the relief activities. The development of VGI and the studies of big data have provided the prerequisite for the introduction of the crowdsourcing service into the GIS. In this research, we tried to solve some of the problems and challenges of crisis management by relying on the characteristics of the crowdsourcing service and its integration with the GIS. Dividing complex emergency processes into atomic crowdsourcing services in a wireless mobile environment is regarded as an appropriate solution. Proposed atomic crowdsourcing services include an informative, content, and confirmative segment. Composing such services can manage sophisticated crisis fields. But, the use of services necessitates a robust and operational framework. This framework must also be capable of simulating the environment in addition to executing, managing, and communicating with the environment. However, current crisis management frameworks cannot cover all these issues at the same time. The present study aims to design a crisis crowdsourcing mathematical framework, which has been constructed upon vector space elements, tensors, and geometrical products. Vectors simulate crisis environment for main objects in a multi-dimensional space. In addition, operations are coincided with possible actions among crisis objects. Framework introduces tools to compose and arrange services in optimum mode to conquer complex aiding process. Since low-speed and high processing load are among the fundamental problems of existing software frameworks, the use of tensors and matrices in frameworks increase the speed and decrease the processing load due to cumulative integrated computing. Further, it offers a mathematical intelligible documentation method, which is able to record the processing details as well as the object status in space. On the other hand, although tensors are widely used in other fields of engineering, there is little record of their use in GIS or crisis management. Thus, with the designed framework, it is possible to use the experiences gained from other disciplines. Different scenarios have confirmed the accuracy and efficiency of the framework in deploying potential of crowds. It offers an environment to easily incorporate people in a wireless mobile application, to tackle emergency process. Indeed, rather than the social aspects of research in providing a new IT-based methodology for employing crowd’s ability, this research has implemented a special math framework for managing new proposed crowdsourcing services that is easily implemented through a mobile GIS system in a wireless network.

A hybrid particle volume-of-fluid method for curvature estimation in multiphase flows

We present a particle method for estimating the curvature of interfaces in volume-of-fluid simulations of multiphase flows. The method is well suited for under-resolved interfaces, and it is shown to be more accurate than the parabolic fitting that is employed in such cases. The curvature is computed from the equilibrium positions of particles constrained to circular arcs and attracted to the interface. The proposed particle method is combined with the method of height functions at higher resolutions, and it is shown to outperform the current combinations of height functions and parabolic fitting. The algorithm is conceptually simple and straightforward to implement on new and existing software frameworks for multiphase flow simulations thus enhancing their capabilities in challenging flow problems. We evaluate the proposed hybrid method on a number of two- and three-dimensional benchmark flow problems and illustrate its capabilities on simulations of flows involving bubble coalescence and turbulent multiphase flows.

OpenFPM: A scalable open framework for particle and particle-mesh codes on parallel computers

Scalable and efficient numerical simulations continue to gain importance, as computation is firmly established as the third pillar of discovery, alongside theory and experiment. Meanwhile, the performance of computing hardware grows through increasingly heterogeneous parallelism, enabling simulations of ever more complex models. However, efficiently implementing scalable codes on heterogeneous, distributed hardware systems becomes the bottleneck. This bottleneck can be alleviated by intermediate software layers that provide higher-level abstractions closer to the problem domain, reducing development times and allowing computational scientists to focus. Here, we present OpenFPM, an open and scalable framework that provides an abstraction layer for numerical simulations using particles and/or meshes. OpenFPM provides transparent and scalable infrastructure for shared-memory and distributed-memory implementations of particles-only and hybrid particle-mesh simulations of both discrete and continuous models, as well as non-simulation codes. This infrastructure is complemented with frequently used numerical routines, as well as interfaces to third-party libraries. We present the architecture and design of OpenFPM, detail the underlying abstractions, and benchmark the framework in applications ranging from Smoothed-Particle Hydrodynamics (SPH) to Molecular Dynamics (MD), Discrete Element Methods (DEM), Vortex Methods, stencil codes (finite differences), and high-dimensional Monte Carlo sampling (CMA-ES), comparing it to the current state of the art and to existing software frameworks. Program summaryProgram Title: OpenFPMProgram Files doi:http://dx.doi.org/10.17632/4yrp8nbm7c.1Licensing provisions: GPLv3Programming language: C++Nature of problem: Writing numerical simulation programs that use meshes, particles, or any combination of the two typically requires long development times, in particular if the code is to scale efficiently on parallel distributed-memory computers. The long development times incur high financial and project-time costs and often lead to sub-optimal program performance as shortcuts are taken. Yet, a large portion of the functionality is common across programs and could be automated or provided as reusable software components, leading to large savings in project costs and potentially improved software performance.Solution method: OpenFPM provides a scalable, highly efficient software platform for numerical simulations using meshes, particles, or any combination of the two on parallel computers. It is based on a well-known set of abstract data types and operators that suffice to express any such simulation, regardless of the application domain. OpenFPM provides reusable, tested, and internally parallelized software components that reduce development times and make parallel computing accessible to computational scientists without extensive knowledge in parallel programming.Additional comments including restrictions and unusual features: OpenFPM is a software library based on which users can implement their simulation codes at a fraction of the development cost. All parallelization and memory handling is transparently done by the library. As its main innovation, OpenFPM makes use of C++ Template Meta Programming in order to enable simulations in arbitrary-dimensional spaces, distribution of arbitrary user-defined C++ objects, and compile-time code optimization and targeting for specific hardware platforms. OpenFPM-based simulations can directly output VTK files for visualization of results and HDF5 files for data archiving.

BindsNET: A Machine Learning-Oriented Spiking Neural Networks Library in Python.

The development of spiking neural network simulation software is a critical component enabling the modeling of neural systems and the development of biologically inspired algorithms. Existing software frameworks support a wide range of neural functionality, software abstraction levels, and hardware devices, yet are typically not suitable for rapid prototyping or application to problems in the domain of machine learning. In this paper, we describe a new Python package for the simulation of spiking neural networks, specifically geared toward machine learning and reinforcement learning. Our software, called BindsNET1, enables rapid building and simulation of spiking networks and features user-friendly, concise syntax. BindsNET is built on the PyTorch deep neural networks library, facilitating the implementation of spiking neural networks on fast CPU and GPU computational platforms. Moreover, the BindsNET framework can be adjusted to utilize other existing computing and hardware backends; e.g., TensorFlow and SpiNNaker. We provide an interface with the OpenAI gym library, allowing for training and evaluation of spiking networks on reinforcement learning environments. We argue that this package facilitates the use of spiking networks for large-scale machine learning problems and show some simple examples by using BindsNET in practice.

Elemental damping formulation: an alternative modelling of inherent damping in nonlinear dynamic analysis

To date, nonlinear dynamic analysis for seismic engineering predominantly employs the classical Rayleigh damping model and its variations. Though earlier studies have identified issues with the use of this model in nonlinear seismic analysis, it still remains the popular choice for engineers as well as for software providers. In this paper a new approach to modelling damping is initiated by formulating the damping matrix at an elemental level. To this regard, two new elemental level discrete damping models adapted from their global counterparts are proposed for its application in nonlinear dynamic analysis. Implementation schemes for these newly proposed models using Newmark incremental method and revised Newmark total equilibrium method is outlined. The performance of these proposed models, compared to existing models, is illustrated by conducting nonlinear dynamic analyses on a four story RC frame designed to Eurocodes. The incremental dynamic analysis study presented in the paper illustrates the fact that both the proposed models seem to produce more reliable results from an engineering perspective in comparison to the global models. It is also shown that the proposed elemental damping models lead to smaller and more realistic damping moments in the plastic hinges. Furthermore, these models could be easily included in existing software frameworks without adding noticeably to the computational effort. The computation time required for these models is approximately equivalent to the one required when using the tangent Rayleigh damping matrix with constant coefficients.

On an infrastructure to support sharing and aggregating pre- and post-publication systems biology research data

The move towards in silico experimentation has resulted in the use of computational models, in addition to traditional experimental models, to generate the raw data that is analysed and published as research findings. This change requires new methods to be introduced to facilitate independent validation of the underlying models and the reported results. The promotion of co-operative research has the potential to help to both validate results and explore wider problem areas. In this paper we leverage and extend two existing software frameworks to develop an infrastructure that has the potential to both promote the sharing of data between researchers pre-publication and enable access to the data for interested parties post-publication. The pre-publication sharing of data would enable larger problem spaces to be explored by distributed research groups; enabling access to the data post-publication would allow reviewers and the wider community to independently verify the published results which would, in the longer term, help to increase confidence in published results. The framework is used to perform reproducible and numerically validated individual-based computational experiments into the onset of colorectal cancer. Existing results are verified and new insights into the top-down versus bottom-up hypothesis of colorectal crypt invasion are given.

RAVE—A Detector-Independent Toolkit to Reconstruct Vertices

A detector-independent toolkit for vertex reconstruction (RAVE = “Reconstruction (of vertices) in Abstract, Versatile Environments”) is presented that allows geometric and kinematic reconstruction of vertices. Both linear and adaptive estimation techniques are covered. Non-Gaussian input data can be handled via the Gaussian-sum technique. Kinematic constraints are taken into account via the Lagrangian formalism. Finally, the toolkit also contains a simple flavor-tagger. Main design goals are ease of use, flexibility for embedding into existing software frameworks, extensibility, and openness. The implementation is based on modern object-oriented techniques, is coded in C++ with interfaces for Java and Python, and follows an open-source approach.

Partnering effects on user–developer conflict and role ambiguity in information system projects

Context Information system development (ISD) has been plagued with high failure rates. This is partially due to the activities being a combination of both a technical and social processes involving stakeholders with conflicting interests. Objective Existing software risk management theories and frameworks offer limited suggestions for actions that can be taken to reduce the chance of failure of ISD projects. Our objective is to examine the connections among some of the more important user related risks in order to shed light on how specific strategies enhance the chance of project success. Method We conducted a sample of information systems project managers to test a multivariate model to explain the impact of pursuing a partnership with users on the conflicts that arise between users and developers, role ambiguity, and subsequent impact on project performance. Results The proposed model was supported, suggesting that user–developer conflict and role ambiguity have a negative impact on performance estimation difficulty, which negatively affects project performance. Conclusion Pursuit of project partnering yields a number of significant relationships in the model indicating an organization can implement practices that reduce risks associated with role ambiguity and conflict in system development projects.

Platforms and testbeds for experimental evaluation of cognitive ad hoc networks

This article reviews, discusses, and classifies the existing experimental platforms and testbeds for experimental evaluation of CR ad hoc networks. The article first describes the internal organization of typical existing software defined radio platforms. Then the existing software frameworks are discussed and classified. Consequently, the challenges involved in integration of these platforms to form multihop networks are discussed, and the major efforts for large-scale testbed implementations described. Finally, limitations of existing platforms and future research challenges toward the design of accurate open testbeds that will enable sound experimental evaluation of CR networks are discussed.

BRAHMS: Novel middleware for integrated systems computation

Biological computational modellers are becoming increasingly interested in building large, eclectic models, including components on many different computational substrates, both biological and non-biological. At the same time, the rise of the philosophy of embodied modelling is generating a need to deploy biological models as controllers for robots in real-world environments. Finally, robotics engineers are beginning to find value in seconding biomimetic control strategies for use on practical robots. Together with the ubiquitous desire to make good on past software development effort, these trends are throwing up new challenges of intellectual and technological integration (for example across scales, across disciplines, and even across time) – challenges that are unmet by existing software frameworks. Here, we outline these challenges in detail, and go on to describe a newly developed software framework, BRAHMS, that meets them. BRAHMS is a tool for integrating computational process modules into a viable, computable system; its generality and flexibility facilitate integration across barriers, such as those described above, in a coherent and effective way. We go on to describe several cases where BRAHMS has been successfully deployed in practical situations. We also show excellent performance in comparison with a monolithic development approach. Additional benefits of developing in the framework include source code self-documentation, automatic coarse-grained parallelisation, cross-language integration, data logging, performance monitoring, and will include dynamic load-balancing and ‘pause and continue’ execution. BRAHMS is built on the nascent, and similarly general purpose, model markup language, SystemML. This will, in future, also facilitate repeatability and accountability (same answers ten years from now), transparent automatic software distribution, and interfacing with other SystemML tools.

The RAVE/VERTIGO vertex reconstruction toolkit and framework

A detector-independent toolkit for vertex reconstruction (RAVE1) is being developed, along with a standalone framework (VERTIGO2) for testing, analyzing and debugging. The core algorithms represent state-of-the-art for geometric vertex finding and fitting by both linear (Kalman filter) and robust estimation methods. Main design goals are ease of use, flexibility for embedding into existing software frameworks, extensibility, and openness. The implementation is based on modern object-oriented techniques, is coded in C++ with interfaces for Java and Python, and follows an open-source approach. A beta release is available.

RAVE—a Detector-independent vertex reconstruction toolkit

A detector-independent toolkit for vertex reconstruction (RAVE 1 1 RAVE = “reconstruction (of vertices) in abstract versatile environments”, ) is being developed, along with a standalone framework (VERTIGO 2 2 VERTIGO = “vertex reconstruction toolkit and interface to generic objects”. ) for testing, analyzing and debugging. The core algorithms represent state of the art for geometric vertex finding and fitting by both linear (Kalman filter) and robust estimation methods. Main design goals are ease of use, flexibility for embedding into existing software frameworks, extensibility, and openness. The implementation is based on modern object-oriented techniques, is coded in C ++ with interfaces for Java and Python, and follows an open-source approach. A beta release is available.

Global computing for bioinformatics.

Global computing, the collaboration of idle PCs via the Internet in a SETI@home style, emerges as a new way of massive parallel multiprocessing with potentially enormous CPU power. Its relations to the broader, fast-moving field of Grid computing are discussed without attempting a review of the latter. This review (i) includes a short table of milestones in global computing history, (ii) lists opportunities global computing offers for bioinformatics, (iii) describes the structure of problems well suited for such an approach, (iv) analyses the anatomy of successful projects and (v) points to existing software frameworks. Finally, an evaluation of the various costs shows that global computing indeed has merit, if the problem to be solved is already coded appropriately and a suitable global computing framework can be found. Then, either significant amounts of computing power can be recruited from the general public, or--if employed in an enterprise-wide Intranet for security reasons--idle desktop PCs can substitute for an expensive dedicated cluster.