A development cycle for automated self-exploration of robot behaviors

JMD Editorial Board

Model-driven and component-based software development, using the UML, has become one of the dominant development paradigms, particularly in business and web application engineering. Unfortunately, model-driven and UML-based development methods are still inferior to conventional software development approaches when it comes to component-based embedded system development. One important aspect is the heterogeneity of embedded systems: they contain both, hardware and software components. Although, component-based development in embedded systems (with hardware components) has a long tradition, there is still a problem of combining it with component-based software development. One reason is the inability of contemporary component technologies to cope with the specific non-functional requirements of embedded systems (e.g., timing, resource consumption). Thus, the major question is how both approaches can be successfully combined.The goal of this chapter is to discuss the problems of embedded systems engineering in the context of a component-based development approach, and to identify specific requirements for a development process under this paradigm. In addition, the chapter proposes an approach to specify software and hardware components in a uniform way, concerning their functional and non-functional properties, so that they can be applied in embedded system development. The method proposed is not yet solving all the problems associated with component-based embedded systems development, but it addresses important issues like hardware/software integration, and how timing and resource issues can be dealt with.KeywordsUnify Modeling LanguageEmbed SystemHardware ComponentDevice DriverUnify Modeling Language DiagramThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Surface transportation has evolved through technology advancements using parallel knowledge areas such as machine learning (ML). However, the transportation industry has not yet taken full advantage of ML. To evaluate this gap, we utilized a literature review approach to locate, categorize, and synthesize the principal concepts of research papers regarding surface transportation systems using ML algorithms, and we then decomposed them into their fundamental elements. We explored more than 100 articles, literature review papers, and books. The results show that 74% of the papers concentrate on forecasting, while multilayer perceptions, long short-term memory, random forest, supporting vector machine, XGBoost, and deep convolutional neural networks are the most preferred ML algorithms. However, sophisticated ML algorithms have been minimally used. The root-cause analysis revealed a lack of effective collaboration between the ML and transportation experts, resulting in the most accessible transportation applications being used as a case study to test or enhance a given ML algorithm and not necessarily to enhance a mobility or safety issue. Additionally, the transportation community does not define transportation issues clearly and does not provide publicly available transportation datasets. The transportation sector must offer an open-source platform to showcase the sector’s concerns and build spatiotemporal datasets for ML experts to accelerate technology advancements.

Contributors

The opacity of machine learning (ML) algorithms is a significant concern in academic and regulatory circles. An emergent sociology of algorithms, however, argues that far from opacity being an inherent quality of algorithms, it is socially constructed and contingent upon certain choices and decisions. In this article, we show that a valorization of opacity is a key component of the epistemic culture of ML experts. While earlier campaigns for mechanical objectivity contrasted the inconsistency of human experts with the reliability of procedures and machines, we found that ML experts valorize precisely those moments when complex algorithms 'surprised' them with unexpected outcomes. They thereby endowed machines with a mysterious capacity to make predictions based on calculations and factors that humans cannot grasp. In this way, they turned opacity from a problem into an epistemic virtue. We trace this valorization of opacity to the jurisdictional struggles through which ML expertise emerged and differentiated itself from its two competitors: the 'expert systems' type of the 'artificial intelligence' sub-field of computer science on the one hand and inferential statistics on the other. In the course of these struggles, ML experts absorbed a theory of human expertise as tacit and inarticulable, extended it to include algorithms, and then leveraged this newly acquired version of opacity to dramatize the differences that separated them from statisticians. The analysis is based on sixty in-depth, semi-structured, and open-ended interviews with ML experts and data scientists working today, as well as historical research on the origins of data science.

Machine Teaching (MT) is an emerging practice where people, without Machine Learning (ML) expertise, provide rich information beyond labels in order to create ML models. MT promises to lower the barrier of entry to creating ML models by requiring a softer set of skills from users than having ML expertise. In this paper, we explore and show how end-users without MT experience successfully build ML models using the MT process, and achieve results not far behind those of MT experts. We do this by conducting two studies. We first investigated how MT experts build models, from which we extracted expert teaching patterns. In our second study, we observed end-users without MT experience create ML models with and without guidance from expert patterns. We found that all users built models comparable to those built by MT experts. Further, we observed that users who received guidance perceived the task to require less effort and felt less mental demand than those who did not receive guidance.

Anomaly detection in laser powder bed fusion using machine learning: A review

The traditional software development model commonly named “waterfall” is unable to cope with the increasing functionality and complexity of modern embedded systems. In addition, it is unable to support the ability for businesses to quickly respond to new market opportunities due to changing requirements. As a response, the software development community developed the Agile Methodologies (e.g., extreme Programming, Scrum) which were also adopted by the Embedded System community. However, failures and bad experiences in applying Agile Methodologies to the development of embedded systems have not been reported in the literature. Therefore, this paper contributes a detailed account of our first-time experiences adopting an agile approach in the prototype development of a wireless environment data acquisition system in an academic environment. We successfully applied a subset of the extreme Programming (XP) methodology to our software development using the Python programming language, an experience that demonstrated its benefits in shaping the design of the software and also increasing productivity. We used an incremental development approach for the hardware components and adopted a “cumulative testing” approach. For the overall development process management, however, we concluded that the Promise/Commitment-Based Project Management (PB-PM/CBPM) was better suited. We discovered that software and hardware components of embedded systems are best developed in parallel or near-parallel. We learned that software components that pass automated tests may not survive in the tests against the hardware. Throughout this rapid prototyping effort, factors like team size and our availability as graduate students were major obstacles to fully apply the XP methodology.

Robot systems are developed using various hardware and software components. In conventional design methodology, each hardware component and its software are strongly coupled such that it is not easy to replace or expand them separately. For the independent development and replacement of hardware and software components, this paper proposes a novel robot development methodology based on the abstractions of software and hardware components in a multilayer architecture for cyber-physical robot systems which conjoin computational and physical resources. We introduce a context generator and a behavior translator for the abstractions in the multilayer architecture. The context generator converts sensory value data into contexts using context scripts. The behavior translator converts a behavior, selected by a software agent that is a computer program deciding an action of a robot, into a sequence of actuator commands for robots using behavior scripts. These together enable two levels of abstraction in which software and hardware components can be developed independently of each other. As a result, software agents can work flawlessly even if hardware components are replaced and vice versa. The effectiveness and applicability of the proposed methodology are demonstrated through experiments, and the related applications are provided.

This paper explores the design of robot systems to take advantage of non-linear dynamic systems models, specifically symmetry breaking phenomena, to self-organize in response to task and environment demands. Recent research in the design of robotics systems has stressed modular, adaptable systems operating under decentralized and distributed control architectures. Cooperative and emergent behavioral structures can be built on these modules by exploiting various forms of communication and negotiation strategies. We focus on the design of individual modules and their cooperative interaction. We draw on nonlinear dynamic system models of human and animal behavior to motivate issues in the design of robot modules and systems. Sonar sensing systems comprising a ring of sonar sensors are used to illustrate the ideas within a networked robotics context, where distributed sensing modules located on multiple robots can interact cooperatively to scan an environment.© (1997) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Human-centered approaches to machine learning have established theoretical foundations, design principles and interaction techniques to facilitate end-user interaction with machine learning systems. Yet, general-purpose toolkits supporting the design of interactive machine learning systems are still missing, despite their potential to foster reuse, appropriation and collaboration between different stakeholders including developers, machine learning experts, designers and end users. In this paper, we present an architectural model for toolkits dedicated to the design of human interactions with machine learning. The architecture is built upon a modular collection of interactive components that can be composed to build interactive machine learning workflows, using reactive pipelines and composable user interfaces. We introduce Marcelle, a toolkit for the design of human interactions with machine learning that implements this model. We illustrate Marcelle with two implemented case studies: (1) a HCI researcher conducts user studies to understand novice interaction with machine learning, and (2) a machine learning expert and a clinician collaborate to develop a skin cancer diagnosis system. Finally, we discuss our experience with the toolkit, along with its limitation and perspectives.

This paper presents a novel integration of Systems Engineering (SE) methodologies and Industry 4.0 (I4.0) technologies in the design of robotic systems, focusing on enhancing underwater robotic missions. Using the conceptual design of an underwater exploration vehicle as a case study, we demonstrate how SE can systematically incorporate I4.0 tools to improve mission performance and meet stakeholder expectations. The study begins with an overview of the SE approach, emphasizing the conceptual design stage and aligning it with the application and case study of design theories. We then explore various I4.0 technologies, highlighting their functional benefits rather than technical specifics and addressing design methods for I4.0. Remotely Operated Vehicles (ROVs) are examined in terms of classification, components, and tasks, showcasing their evolution driven by technological advancements, thus tackling the complexity and design of complex systems. The core of our study involves defining stakeholder expectations, using quality function deployment for requirements definition, and performing a functional and logical decomposition of the ROV system. To deal with design fixation within the design team, we developed a tool to help integrate new technologies by also empathizing with their functional capabilities rather than the technology itself. Our approach underscores the importance of understanding and incorporating new technologies functionally, aligning with the transition towards Industry/Society 5.0. This work not only illustrates the synergy between SE and I4.0, but also offers a structured methodology for advancing the design and functionality of complex systems, setting a blueprint for future developments in this field.

As smart and automated applications pervade our lives, an increasing number of software developers are required to incorporate machine learning (ML) techniques into application development. However, acquiring the ML skill set can be nontrivial for software developers owing to both the breadth and depth of the ML domain. Aims: We seek to understand the challenges developers face in the process of ML application development and offer insights to simplify the process. Despite its importance, there has been little research on this topic. A few existing studies on development challenges with ML are outdated, small scale, or they do no involve a representative set of developers. Method: We conduct an empirical study of ML-related developer posts on Stack Overflow. We perform in-depth quantitative and qualitative analyses focusing on a series of research questions related to the challenges of developing ML applications and the directions to address them. Results: Our findings include: (1) ML questions suffer from a much higher percentage of unanswered questions on Stack Overflow than other domains; (2) there is a lack of ML experts in the Stack Overflow QA community; (3) the data preprocessing and model deployment phases are where most of the challenges lay; and (4) addressing most of these challenges require more ML implementation knowledge than ML conceptual knowledge. Conclusions: Our findings suggest that most challenges are under the data preparation and model deployment phases, i.e., early and late stages. Also, the implementation aspect of ML shows much higher difficulty level among developers than the conceptual aspect.

Authors

About The Author