Papers by Janos Sztipanovits
2021 6th International Conference on Machine Learning Technologies, 2021
Machine learning models have prevalent applications in many real-world problems, which increases ... more Machine learning models have prevalent applications in many real-world problems, which increases the importance of correctness in the behaviour of these trained models. Finding a good test case that can reveal the potential failure in these trained systems can help to retrain these models to increase the correctness. For a well-trained model, the occurrence of a failure is rare. Consequently, searching these rare scenarios by evaluating each sample in input search space or randomized search would be costly and sometimes intractable due to large search space, limited computational resources, and available time. In this paper, we tried to address this challenge of finding these failure scenarios faster than traditional randomized search. The central idea of our approach is to separate the input data space in region of high failure probability and region of low/minimal failure probability based on the observation made by training data, data drawn from realworld statistics, and knowledge from a domain expert. Using these information, we can design a generative model from which we can generate scenarios that have a high likelihood to reveal the potential failure. We evaluated this approach on two different experimental scenarios and able to speed up the discovery of such failures a thousand-fold faster than the traditional randomized search. Source code of experimentation can be found at here.
TE-SAT: Transactive Energy Simulation and Analysis Toolsuite
2020 IEEE Workshop on Design Automation for CPS and IoT (DESTION)
Transactive Energy (TE) is an emerging discipline that utilizes economic and control techniques f... more Transactive Energy (TE) is an emerging discipline that utilizes economic and control techniques for operating and managing the power grid effectively. Distributed Energy Resources (DERs) represent a fundamental shift away from traditionally centrally managed energy generation and storage to one that is rather distributed. However, integrating and managing DERs into the power grid is highly challenging owing to the TE implementation issues such as privacy, equity, efficiency, reliability, and security. The TE market structures allow utilities to transact (i.e., buy and sell) power services (production, distribution, and storage) from/to DER providers integrated as part of the grid. Flexible power pricing in TE enables power services transactions to dynamically adjust power generation and storage in a way that continuously balances power supply and demand as well as minimize cost of grid operations. Therefore, it has become important to analyze various market models utilized in different TE applications for their impact on above implementation issues.In this demo, we show-case the Transactive Energy Simulation and Analysis Toolsuite (TE-SAT) with its three publicly available design studios for experimenting with TE markets. All three design studios are built using metamodeling tool called the Web-based Graphical Modeling Environment (WebGME). Using a Git-like storage and tracking backend server, WebGME enables multi-user editing on models and experiments using simply a web-browser. This directly facilitates collaboration among different TE stakeholders for developing and analyzing grid operations and market models. Additionally, these design studios provide an integrated and scalable cloud backend for running corresponding simulation experiments.
C2WT-TE: A model-based open platform for integrated simulations of transactive smart grids
2016 Workshop on Modeling and Simulation of Cyber-Physical Energy Systems (MSCPES), 2016
Evaluation of the smart grid in the presence of dynamic market-based pricing and complex networks... more Evaluation of the smart grid in the presence of dynamic market-based pricing and complex networks of small and large producers, consumers, and distributers is a very difficult task. It not only involves multiple, interacting, heterogeneous cyber-physical domains, but also requires tight integration of power markets, dynamic pricing and transactions, and price-sensitive consumer behavior, where consumers can also be producers of power. These dynamics introduce a huge challenge of maintaining stability of the power grid. Moreover, evolving business models and regulatory environment, including human factors need to be a key part of the evaluation as they directly affect the demand-response in the grid. Furthermore, as sensors and computations are becoming more distributed on edge devices and as they often employ normal communication channels, cyber security of the critical power grid infrastructure has become ever more important to prevent cyber intrusions and attack. Current research has largely focused on one or a few of these challenges and the simulation tools developed cater to these individual tasks, such as network simulators, power distribution simulators, or human or organization and policy simulators. However, to attain a deeper understanding of the transactive grid behavior, analysis of grid stability, and to optimize resources both from consumers', and generators' perspective, a comprehensive platform is needed to facilitate end-to-end evaluation of all of these aspects. This paper describes an open platform that provides a coherent framework for integrated transactive energy simulations of smart grids and is readily customizable and extensible for various simulation tools.
Design and implementation of an execution environment for knowledge-based systems
The knowledge-based synthesis and dynamic modification of large processing configurations consist... more The knowledge-based synthesis and dynamic modification of large processing configurations consisting both symbolic and numerical computing modules is a common problem in large-scale engineering applications. The goal of the work described in this thesis was to develop an execution environment which supports this process on various computer architectures, by providing the same parallel computational model, regardless of the underlying hardware system. First, an overview of the relevant research areas is given, which covers the topic of distributed problem solving, parallel computational models for knowledge-based systems (Actors, etc.) and the issues in coupling symbolic and numerical computations. Next, the data-flow computational model, and various fine- and large-grain data-flow architectures are discussed. A macro-dataflow execution environment was developed for the knowledge-based integration of complex engineering application programs. The last part of the thesis contains two application examples for the execution environment which include a parallel-rule interpreter capable of non-monotonic reasoning, and a knowledge-based experiment designer and operator interface for medical instrumentation.
Towards Bridging the Gap Between Model- and Data- Driven Tool Suites for Cyber-Physical Systems
2019 IEEE/ACM 5th International Workshop on Software Engineering for Smart Cyber-Physical Systems (SEsCPS), 2019
Model-driven approaches in developing and operating Cyber-Physical Systems are increasingly compl... more Model-driven approaches in developing and operating Cyber-Physical Systems are increasingly complemented by data-driven methods. Examples for their use cases are the analysis of model repositories for discovering patterns and relationships in models, the design-time learning of approximate system and environment models from data, and the detection of divergence from design-time assumptions during operations. In this paper we argue that model-and data-driven approaches have combined use cases that that need complementary services provided by modeling and data analytic frameworks. However, convergence of model-driven and data-driven methods is hindered by the strongly different tool infrastructure. The paper summarizes the integration challenges and proposes a semantic bridge as a solution for filling the gap between the model - and data-driven tool suites.
Cyber‐Physical Vulnerability Analysis of IoT Applications Using Multi‐Modeling
Modeling and Design of Secure Internet of Things, 2020
Semantic Integration Platform for Cyber-Physical System Design
2019 Design, Automation & Test in Europe Conference & Exhibition (DATE), 2019
Cyber-Physical Systems (CPS) are establishing heterogeneous engineering domains leading to engine... more Cyber-Physical Systems (CPS) are establishing heterogeneous engineering domains leading to engineering processes that span multiple design disciplines with separate modeling approaches, design flows and supporting tool suites. One of the challenges of design automation in CPS is the deep integration of models, tools and design flows such that design trade-offs across traditionally isolated design disciplines is facilitated. In this paper we overview experience and results gained along the implementation of an experimental design automation tool suite, OpenMETA, created for a complex CPS design challenge in the ground vehicle domain. The focus of the paper is domain agnostic methods and tools providing infrastructure for the model- and tool- integration platforms in OpenMETA. We present the arguments leading to the creation of the integration platforms instead of pursuing ad-hoc integration of heterogeneous tools and provide details on facilitating semantic integration.
Graphical programming for the transputer
[1990] Proceedings. The Twenty-Second Southeastern Symposium on System Theory
The Multigraph programming environment provides a very-high-level programmer interface for the de... more The Multigraph programming environment provides a very-high-level programmer interface for the development of parallel and real-time processing systems. It is specifically targeted for large systems wishing to integrate a knowledge-based synthesis technique with standard numerical techniques. The result is a graphical editing environment where the user models the structure of the desired computation. Subsequently, symbolic techniques are used to translate
Journal of Responsible Innovation, 2021
This study develops a comparative, sociotechnical design perspective for interdisciplinary teams ... more This study develops a comparative, sociotechnical design perspective for interdisciplinary teams of social scientists and computer scientists. Sociotechnical design refers to identifying both technical and governance challenges and to understanding the ways in which the two types of problems affect and define each other. Approaching design as an open-ended, iterative process, the study develops a triple comparative perspective to problem finding and solutions: across two types of technological systems (the smart grid and connected and automated vehicles), three areas of societal implication and values (safety, equity, and privacy), and two continents (North America and Europe with a focus on the U.S. and Germany). The study then describes the implementation in an international collaboration of research and teaching. The collaborative experience and comparative research provide insights into the salience of the values across technological systems, portability of solutions across technological systems, and potential for policy harmonization across countries.
SURE: an experimentation and evaluation testbed for CPS security and resilience: demo abstract
International Conference on Cyber-Physical Systems, Apr 11, 2016
In-depth consideration and evaluation of security and resilience is necessary for developing the ... more In-depth consideration and evaluation of security and resilience is necessary for developing the scientific foundations and technology of Cyber-Physical Systems (CPS). In this demonstration, we present SURE [1], a CPS experimentation and evaluation testbed for security and resilience focusing on transportation networks. The testbed includes (1) a heterogeneous modeling and simulation integration platform, (2) a Web-based tool for modeling CPS in adversarial environments, and (3) a framework for evaluating resilience using attacker-defender games. Users such as CPS designers and operators can interact with the testbed to evaluate monitoring and control schemes that include sensor placement and traffic signal configuration.
Springer eBooks, 2003
Self-adaptive software systems use observations of their own behavior, and that of their environm... more Self-adaptive software systems use observations of their own behavior, and that of their environment, to select and enact adaptations in accordance with some objective(s). This adaptation is a higher-level system function that performs optimizations, manages faults, or otherwise supports achieving an objective via changes in the running system. In this paper, we show how this capability can be realized using techniques found in hierarchical control systems, and we discuss interrelated issues of stability, assurance, and implementation.
arXiv (Cornell University), Feb 18, 2023
Problem Formulation and Approach: AI has proven its potential in system design and operation [1,2... more Problem Formulation and Approach: AI has proven its potential in system design and operation [1,2,3]. In this work, we want to leverage AI for finding a universal minimum drag hull shape for AUV. First, we formulate the AUV hull design problem as an optimization problem. Let the hull shape indicated as Λ is defined using a multivariate parameter , and is the function that maps a given 3D shape Λ with a drag force (_) (: Λ → _). If is the design space of search, the optimization problem can be formulated as Λ * = () ∈ The optimization involves running the CFD analysis on chosen hull shape. Since running these CFD simulations are very time-consuming and computationally costly, we chose AI-based Bayesian optimization [4,5,6,7] as our optimization algorithm in a loop with CFD analysis to find the optimal design. Second, we want to study/search for a universal optimal design that is near optimal across all environmental and operating conditions. For creating the different environment and operating conditions for AUV, we consider five different environmental conditions (turbulence intensity of flow {0.1,2,5,10, 20} percent of mean flow velocity) and five different operating conditions (velocity of AUV {1,2.5,5,7.5,10} in meters/second). The operating velocity and turbulence intensity values are taken from empirical ranges of underwater vehicles' operation. Accordingly, the cartesian product of these two sets creates 25 different ABSTRACT In Autonomous Underwater Vehicles (AUVs) design, hull resistance is an important factor in determining the power requirements and range of vehicle and consequently affect battery size, weight, and volume requirement of the design. In this paper, we leverage on AI based optimization algorithm along with Computational Fluid Dynamics (CFD) simulation to study the optimal hull design that minimizing the resistance. By running the CFD based optimization at different operating velocity and turbulence intensity, we want to study/search the possibility of a universal design that will provide least resistance/ near optimal design across all operating condition (operating velocity) and environmental conditions (turbulence intensity). Early result demonstrated that the optimal design found at low velocity and low turbulence condition performs very poor at high velocity and high turbulence conditions. However, design that is optimal at high velocity and high turbulence condition performs near optimal across many considered velocity and turbulence conditions.
arXiv (Cornell University), May 24, 2018
The Behavior-Interaction-Priority (BIP) framework-rooted in rigorous semantics-allows the constru... more The Behavior-Interaction-Priority (BIP) framework-rooted in rigorous semantics-allows the construction of systems that are correct-by-design. BIP has been effectively used for the construction and analysis of large systems such as robot controllers and satellite on-board software. Nevertheless, the specification of BIP models is done in a purely textual manner without any code editor support. To facilitate the specification of BIP models, we present DesignBIP, a web-based, collaborative, version-controlled design studio. To promote model scaling and reusability of BIP models, we use a graphical language for modeling parameterized BIP models with rigorous semantics. We present the various services provided by the design studio, including model editors, code editors, consistency checking mechanisms, code generators, and integration with the JavaBIP tool-set.
Linköping electronic conference proceedings, Feb 26, 2019
Modelica has been around as a language from the late 1990's and since then a range of compilers a... more Modelica has been around as a language from the late 1990's and since then a range of compilers and editors have emerged. Currently none of these environments provide a web-based user interface and follow the approach of requiring each end-user to install the application (typically together with a set of dependencies) on their local machine. This in itself may or may not be of major concern. Of more importance is their current lack of a seamless collaborative approach to modeling. This paper presents the first web-based collaborative graphical and textual modeling environment for Modelica based on WebGME and OpenModelica. Graphical composition of Modelica models from component libraries is supported via WebGME. Textual editing of the composite model is possible via OMWebBook.
Compositional and incremental modeling and analysis for high-confidence distributed embedded control systems
Automatic underwater vehicle hull Design optimization is a complex engineering process for genera... more Automatic underwater vehicle hull Design optimization is a complex engineering process for generating a UUV hull with optimized properties on a given requirement. First, it involves the integration of involved computationally complex engineering simulation tools. Second, it needs integration of a sample efficient optimization framework with the integrated toolchain. To this end, we integrated the CAD tool called FreeCAD with CFD tool openFoam for automatic design evaluation. For optimization, we chose Bayesian optimization (BO), which is a well-known technique developed for optimizing time-consuming expensive engineering simulations and has proven to be very sample efficient in a variety of problems, including hyper-parameter tuning and experimental design. During the optimization process, we can handle infeasible design as constraints integrated into the optimization process. By integrating domain-specific toolchain with AI-based optimization, we executed the automatic design optimization of underwater vehicle hull design. For empirical evaluation, we took two different use cases of real-world underwater vehicle design to validate the execution of our tool.
Transportation Networks
Springer eBooks, May 30, 2018
Cyber-physical systems (CPS) are smart systems that include engineered interacting networks of ph... more Cyber-physical systems (CPS) are smart systems that include engineered interacting networks of physical and computational components. The tight integration of a wide range of heterogeneous components enables new functionality and quality of life improvements in critical infrastructures such as smart cities, intelligent buildings, and smart energy systems. One approach to study CPS uses both simulations and hardware-in-theloop (HIL) to test the physical dynamics of hardware in a controlled environment. However, because CPS experiment design may involve domain experts from multiple disciplines who use different simulation tool suites, it can be a challenge to integrate the heterogeneous simulation languages and hardware interfaces into a single experiment. The National Institute of Standards and Technology (NIST) is working on the development of a universal CPS environment for federation (UCEF) that can be used to design and run experiments that incorporate heterogeneous physical and computational resources over a wide geographic area. This development environment uses the High Level Architecture (HLA), which the Department of Defense has advocated for co-simulation in the field of distributed simulations, to enable communication between hardware and different simulation languages such as Simulink and LabVIEW. This paper provides an overview of UCEF and motivates how the environment could be used to develop energy experiments using an illustrative example of an emulated heat pump system.
Uploads
Papers by Janos Sztipanovits