Zoekresultaten
Volg ons op
Volg ons op
Systems Engineering (SE) is a methodical approach to the development of New (high-tech) products. In 2012 a group of Dutch universities of applied Sciences and companies started a so-called RAAK MBK project to make the Tools of SE more accessible for SME’s. At the same time these tools can be used for multidisciplinary students projects.
DOCUMENT
Additions to the book "Systems Design and Engineering" by Bonnema et.al. Subjects were chosen based on the Systems Engineering needs for Small and Medium Enterprises, as researched in the SESAME project. The
MULTIFILE
The past two years I have conducted an extensive literature and tool review to answer the question: “What should software engineers learn about building production-ready machine learning systems?”. During my research I noted that because the discipline of building production-ready machine learning systems is so new, it is not so easy to get the terminology straight. People write about it from different perspectives and backgrounds and have not yet found each other to join forces. At the same time the field is moving fast and far from mature. My focus on material that is ready to be used with our bachelor level students (applied software engineers, profession-oriented education), helped me to consolidate everything I have found into a body of knowledge for building production-ready machine learning (ML) systems. In this post I will first define the discipline and introduce the terminology for AI engineering and MLOps.
LINK
This chapter discusses how to build production-ready machine learning systems. There are several challenges involved in accomplishing this, each with its specific solutions regarding practices and tool support. The chapter presents those solutions and introduces MLOps (machine learning operations, also called machine learning engineering) as an overarching and integrated approach in which data engineers, data scientists, software engineers, and operations engineers integrate their activities to implement validated machine learning applications managed from initial idea to daily operation in a production environment. This approach combines agile software engineering processes with the machine learning-specific workflow. Following the principles of MLOps is paramount in building high-quality production-ready machine learning systems. The current state of MLOps is discussed in terms of best practices and tool support. The chapter ends by describing future developments that are bound to improve and extend the tool support for implementing an MLOps approach.
LINK
Both Software Engineering and Machine Learning have become recognized disciplines. In this article I analyse the combination of the two: engineering of machine learning applications. I believe the systematic way of working for machine learning applications is at certain points different from traditional (rule-based) software engineering. The question I set out to investigate is “How does software engineering change when we develop machine learning applications”?. This question is not an easy to answer and turns out to be a rather new, with few publications. This article collects what I have found until now.
LINK
Undergraduate students who seek a bachelor degree in Dutch universities of applied sciences are supposed to learn also research skills so that they can provide innovative solutions to real problems of the society and businesses in their future careers. Current education and textbooks on research skills are not tuned well to software engineering disciplines. This paper describes our vision about the scope and model of the research suitable for software engineering disciplines in Dutch universities of applied sciences. Based on literature study we identify a number of research models that are commonly used in computer science. Through reviewing a number of graduation reports in our university, we further identify which of the research models are most suitable for the (graduation) projects of software engineering disciplines and also investigate their shortcomings with respect to the desired research skills. Our study reveals that the approach of most graduation works is close to the implementation-based (also called build-based or proof by example based) research model. In order to be considered as a realization of sound applied research, however, most of theses graduation works need to be improved on a number of aspects such as problem context definition, system/prototype evaluation, and critical literature study.
DOCUMENT
The importance of teaching engineering students innovation development is commonly clearly understood. It is essential to achieve products which are attractive and profitable in the market. To achieve this, an institute of engineering education has to provide students with needed knowledge, skills and attitudes including both technical and business orientation. This is important especially for SME’s. Traditionally, education of engineering provides students with basic understanding how to solve common technical problems. However companies need wider view to achieve new products. Universities of applied Sciences in Oulu and Eindhoven want to research what is the today’s educational situation for this aim, to find criteria to improve the content of the educational system, and to improve the educational system. Important stakeholders are teachers and students within the institute but also key-persons in companies. The research is realized by questionnaires and interviews from which a current situation can be found. The research will also include the opinion of management who give possibilities to change the curriculum. By this research more insight will be presented about how to re-design a current curriculum. The research will act as basis for this discussion in SEFI-conference about formulating a curriculum that includes elements for wide-ranging knowledge and skills to achieve innovations especially in SME’s.
DOCUMENT
Designing privacy-protecting Information Systems (ISs), i.e., realizing the Privacy by Design (PbD) principle, is a challenging task. This challenge stems from having many stakeholders and design trade-offs involved, which cause uncertainty in defining the problem, eliciting soft requirements, and making design trade-offs among many contending objectives. As creating a formal model of such settings is often infeasible, applying a conventional engineering design method alone may not result in elucidating users' needs and/or devising a viable design that is acceptable for all parties (e.g., end-users and data subjects). This contribution aims at enriching engineering approaches for privacy-protecting ISs with the so-called design-thinking approach. Design-thinking, initially used for product and service design, has been applied to the areas where there are interactions among people, organizations and technologies, in order to elucidate user needs and concerns that are insufficiently formulated and/or hidden in tacit knowledge. In this contribution, we elaborate on three main PbD components, namely problem space, solution space and mapping space. We, further, analyze the shortcomings of traditional engineering approaches for privacy protection as well as the potentials and shortcomings of design-thinking in general. Finally, we present our practical experience with applying the design-thinking approach to the problem of PbD for ISs. We foresee the applicability of design-thinking for elucidating the problem space as well as for making design trade-off among contending values in order to come up with a viable design option.
LINK
In the fall of 1999, we started, the Integrated Product Development- Collaborative Engineering ( IPD-CE) project as a first pilot. We experimented with modern communication technology in order to find useful tools for facilitating the cooperative work and the contacts of all the participants. Teams have been formed with engineering students from Lehigh University in the US, the Fontys University in Eindhoven, The Netherlands and from the Otto-von-Guericke University in Magdeburg, Germany. In the fall of 2000 we continued and also cooperated with the Finnish Oulu Polytechnic. It turned out that group cohesion stayed low (students did not meet in real life), and that Internet is not mature enough yet for desktop video conferencing. Chatting and email were in these projects by far the most important communication media. We also found out that the use of a Computer Support for Cooperative Work (CSCW) server is a possibility for information interchange. The server can also be used as an electronic project archive. Points to optimise are: 1. We didn't fully match the complete assignments of the groups; 2. We allowed the groups to divide the work in such parts that those were developed and prototyped almost locally; 3. We haven't guided the fall 2000 teams strong enough along our learning curve and experiences from previous groups. 4. We didn't stick strong enough to the, by the groups developed, protocols for email and chat sessions. 5. We should facilitate video conferencing via V-span during the project to enhance the group performance and commitment.
DOCUMENT
The current set of research methods on ictresearchmethods.nl contains only one research method that refers to machine learning: the “Data analytics” method in the “Lab” strategy. This does not reflect the way of working in ML projects, where Data Analytics is not a method to answer one question but the main goal of the project. For ML projects, the Data Analytics method should be divided in several smaller steps, each becoming a method of its own. In other words, we should treat the Data Analytics (or more appropriate ML engineering) process in the same way the software engineering process is treated in the framework. In the remainder of this post I will briefly discuss each of the existing research methods and how they apply to ML projects. The methods are organized by strategy. In the discussion I will give pointers to relevant tools or literature for ML projects.
LINK
