Models@run.time: a guided tour of the state of the art and research challenges

Bencomo, Nelly; Götz, Sebastian; Song, Hui

doi:10.1007/s10270-018-00712-x

Cited by 117 publications

(80 citation statements)

References 140 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Maintaining a model at runtime not only allows checking of a wider range of properties, but, by allowing changes in the model to be reflected back into the system (not just from the system to the model), we gain the ability to perform model based adaption of the deployed system. Such self-adaptation is common in the models@runtime approach [38].…”

Section: Discussionmentioning

confidence: 99%

BigraphTalk: Verified Design of IoT Applications

Archibald

Shieh

et al. 2020

IEEE Internet Things J.

View full text Add to dashboard Cite

Graphical IoT device management platforms, such as IoTtalk, make it easy to describe interactions between IoT devices. Applications are defined by dragging-and-dropping devices and specifying how they are connected, e.g. a door sensor controlling a light. While this allows simple and rapid development, it remains possible to specify unwanted device configurations -such as using the same device to drive a motor up and down simultaneously, risking damaging the motor.We propose BigraphTalk, a verification framework for IoTtalk that utilizes formal techniques, based on bigraphs, to statically guarantee that unwanted configurations do not arise. In particular, we check for invalid connections between devices, as well as type errors, e.g. passing a float to a boolean switch. To the best of our knowledge, BigraphTalk is the first platform to support the graphical specification of correct-by-design IoT applications.BigraphTalk provides fully automated verification and feedback without end-users ever needing to specify a bigraph. This means any application, specifiable in IoTtalk, is guaranteed, so long as verification succeeds, not to violate the given configuration constraints when deployed; with no extra cost to the user.

show abstract

Section: Discussionmentioning

confidence: 99%

BigraphTalk: Verified Design of IoT Applications

Archibald

Shieh

et al. 2020

IEEE Internet Things J.

View full text Add to dashboard Cite

show abstract

“…Our work uses new evidence found at runtime to (i) provide extra information during decision-making and, (ii) find opportunities to improve the SLA conformance by updating the rewards associated with NFRs and adaptation actions. As future work, we aim to exploit the potential for modularity of the architecture proposed by adding other runtime K models such as goal models [4]. Further, Bayesian learning offers a great potential for self-explanation [14].…”

Section: Discussionmentioning

confidence: 99%

“…Significant advances have been made in applying models at runtime, most notably in adaptive systems [3]. However research challenges still prevail [4]. For example, further techniques to deal with uncertainty [11] and incompleteness of information from systems and their environment are needed for building future software systems [1], [9].…”

Section: Introductionmentioning

confidence: 99%

Knowledge Base K Models to Support Trade-Offs for Self-Adaptation using Markov Processes

Paucar

Bencomo

2019

2019 IEEE 13th International Conference on Self-Adaptive and Self-Organizing Systems (SASO)

Self Cite

View full text Add to dashboard Cite

Runtime models support decision-making and reasoning for self-adaptation based on both design-time knowledge and information that may emerge at runtime. In this paper, we demonstrate a novel use of Partially Observable Markov Decision Processes (POMDPs) as runtime models to support the decision-making of a Self Adaptive System (SAS) in the context of the MAPE-K loop. The trade-off between the non-functional requirements (NFRs) has been embodied as a POMDP in the context of the MAPE-K loop. Using Bayesian learning, the levels of satisficement of the NFRs are inferred and updated during execution in the form of runtime models in the Knowledge Base. We evaluate our work with a case study of the networking application domain.

show abstract

“…-Model analysis the challenge of techniques for analyzing models (e.g., for performance or correctness [13]), along with principles relate to understanding what makes a good model. -Models at runtime the use of models to manage and understand systems after they have been deployed and as they execute behavior [14]. Substantial research has taken place regarding this challenging to identify techniques and tools for automatically reflecting changes from a system into changes in models, and vice versa.…”

Section: Challenges From 2007 Through Present Daymentioning

confidence: 99%

Grand challenges in model-driven engineering: an analysis of the state of the research

et al. 2020

View full text Add to dashboard Cite

In 2017 and 2018, two events were held-in Marburg, Germany, and San Vigilio di Marebbe, Italy, respectively-focusing on an analysis of the state of research, state of practice, and state of the art in model-driven engineering (MDE). The events brought together experts from industry, academia, and the open-source community to assess what has changed in research in MDE over the last 10 years, what challenges remain, and what new challenges have arisen. This article reports on the results of those meetings, and presents a set of grand challenges that emerged from discussions and synthesis. These challenges could lead to research initiatives for the community going forward. Keywords Model-driven engineering • Grand challenge • Research roadmap 1 Introduction The field of model-driven engineering [1] (MDE) has evolved substantially from the earliest work on UML in the 1990s, through to seminal research on metamodeling, model transformation, and model management in the earlyto-mid-2000s. MDE has made incredible contributions to leverage abstraction and automation in almost every area of software and systems development and analysis. In many domains, including railway systems, automotive, business process engineering, and embedded systems, models are key to success in modern software engineering processes. How-Communicated by Bernhard Rumpe.

show abstract

Models@run.time: a guided tour of the state of the art and research challenges

Cited by 117 publications

References 140 publications

BigraphTalk: Verified Design of IoT Applications

BigraphTalk: Verified Design of IoT Applications

Knowledge Base K Models to Support Trade-Offs for Self-Adaptation using Markov Processes

Grand challenges in model-driven engineering: an analysis of the state of the research

Contact Info

Product

Resources

About