Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00034
Cameron Hickert, Ali Tekeoglu, Ryan Watson, Joseph Maurio, Daniel P. Syed, Jeffrey S. Chavis, G. Brown, Tamim I. Sookoor
Incorporating smart technology into critical infrastructure (CI) promises substantial efficiency improvements as networks of machines communicate and make rapid decisions autonomously. Yet the promise of greater efficiency that such cyber-physical systems (CPS) bring is tempered by increased fragility unless machine-to-machine (M2M) trust is enhanced, particularly in Internet-of-Things (IoT) networks. This work makes two contributions toward improving M2M trust. First, it proposes a multifaceted trust framework comprised of identity verification, experience, context, and recommendation scores to enable high-integrity M2M interactions. Second, this trust framework is implemented via an IoT-friendly distributed ledger on a physical testbed, where it is shown to identify and mitigate errors due to a compromised system component. This implementation mirrors real-world IoT systems in which resource- constrained endpoint devices pose trust score computation chal-lenges and the number of devices raises scalability obstacles for information sharing among nodes.
{"title":"Trust Me, I'm Lying: Enhancing Machine-to-Machine Trust","authors":"Cameron Hickert, Ali Tekeoglu, Ryan Watson, Joseph Maurio, Daniel P. Syed, Jeffrey S. Chavis, G. Brown, Tamim I. Sookoor","doi":"10.1109/iccps54341.2022.00034","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00034","url":null,"abstract":"Incorporating smart technology into critical infrastructure (CI) promises substantial efficiency improvements as networks of machines communicate and make rapid decisions autonomously. Yet the promise of greater efficiency that such cyber-physical systems (CPS) bring is tempered by increased fragility unless machine-to-machine (M2M) trust is enhanced, particularly in Internet-of-Things (IoT) networks. This work makes two contributions toward improving M2M trust. First, it proposes a multifaceted trust framework comprised of identity verification, experience, context, and recommendation scores to enable high-integrity M2M interactions. Second, this trust framework is implemented via an IoT-friendly distributed ledger on a physical testbed, where it is shown to identify and mitigate errors due to a compromised system component. This implementation mirrors real-world IoT systems in which resource- constrained endpoint devices pose trust score computation chal-lenges and the number of devices raises scalability obstacles for information sharing among nodes.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125134509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00048
Upinder Kaur, Z. Berkay Celik, R. Voyles
Cyber-Physical Systems (CPS) increasingly use multiple robots as edge devices to enhance their functionalities. However, this introduces new security vulnerabilities such as control channel attacks and false data injection that an adversary can exploit to put the users and environment at risk. In this paper, we build a robust malware detection system strengthened by carefully crafted adversarial samples. We generate adver-sarial samples within the bounds of domain constraints and integrate them into model training to improve the model's robustness. Additionally, we formulate an objective function to distribute the computation of malware detection to multiple edges, making optimal use of the robot mesh network to reduce power consumption. In the adjoining poster, we show the details of the dataset and the models, and illustrate the specifics of our contributions.
{"title":"Robust and Energy Efficient Malware Detection for Robotic Cyber-Physical Systems","authors":"Upinder Kaur, Z. Berkay Celik, R. Voyles","doi":"10.1109/iccps54341.2022.00048","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00048","url":null,"abstract":"Cyber-Physical Systems (CPS) increasingly use multiple robots as edge devices to enhance their functionalities. However, this introduces new security vulnerabilities such as control channel attacks and false data injection that an adversary can exploit to put the users and environment at risk. In this paper, we build a robust malware detection system strengthened by carefully crafted adversarial samples. We generate adver-sarial samples within the bounds of domain constraints and integrate them into model training to improve the model's robustness. Additionally, we formulate an objective function to distribute the computation of malware detection to multiple edges, making optimal use of the robot mesh network to reduce power consumption. In the adjoining poster, we show the details of the dataset and the models, and illustrate the specifics of our contributions.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117109161","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00050
Yushi Ogiwara, Ayanori Yorozu, A. Ohya, H. Kawashima
TF library is a frequently used package in ROS, which manages transformations between coordinate systems as a directed tree structure, and enables registrations and calculation of coordinate transformation information. TF tree access is not scalable due to a giant lock and does not provide the latest data. The proposed method solves these problems by applying the fine-grained locking method and the two phase locing. We show that the proposed method achieves up to 143 times faster throughput, up to 208 times shorter latency, and up to 132 times data freshness than the existing methods.
{"title":"Making ROS TF Transactional","authors":"Yushi Ogiwara, Ayanori Yorozu, A. Ohya, H. Kawashima","doi":"10.1109/iccps54341.2022.00050","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00050","url":null,"abstract":"TF library is a frequently used package in ROS, which manages transformations between coordinate systems as a directed tree structure, and enables registrations and calculation of coordinate transformation information. TF tree access is not scalable due to a giant lock and does not provide the latest data. The proposed method solves these problems by applying the fine-grained locking method and the two phase locing. We show that the proposed method achieves up to 143 times faster throughput, up to 208 times shorter latency, and up to 132 times data freshness than the existing methods.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"2016 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134143197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00038
R. Bhadani, J. Sprinkle, K. L. Head
In this Work-in-Progress poster, we summarize a model- based design approach for a NEMA dual-ring-barrier, eight- phase traffic signal controller that is capable of interfacing with the open-source SUMO software. The goal is to create a model that can support testing and evaluation of advanced cooperative automated driving traffic management and control systems at a fraction of cost of proprietary software.
{"title":"Model-based Design of NEMA-Compliant Dual-ring-barrier Traffic Signal Controller","authors":"R. Bhadani, J. Sprinkle, K. L. Head","doi":"10.1109/iccps54341.2022.00038","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00038","url":null,"abstract":"In this Work-in-Progress poster, we summarize a model- based design approach for a NEMA dual-ring-barrier, eight- phase traffic signal controller that is capable of interfacing with the open-source SUMO software. The goal is to create a model that can support testing and evaluation of advanced cooperative automated driving traffic management and control systems at a fraction of cost of proprietary software.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134518881","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00018
Christian Llanes, Matthew Abate, S. Coogan
We present a runtime assurance (RTA) mechanism for ensuring safety of a controlled dynamical system and an application to collision avoidance of two unmanned aerial vehicles (UAVs). We consider a dynamical system controlled by an unverified and potentially unsafe primary controller that might, e.g., lead to collision. The proposed RTA mechanism computes at each time the reachable set of the system under a backup control law. We then develop a novel optimization problem based on control barrier functions that filters the primary controller when necessary in order to keep the system's reachable set within reach of a known, but conservative, safe region. The theory of mixed monotone systems is leveraged for efficient reachable set computation and to achieve a tractable optimization formulation. We demonstrate the proposed RTA mechanism on a dual multirotor UAV case study which requires a fast controller update rate as a result of the small time-scale rotational dynamics. In implementation, the algorithm computes the reachable set of an eight dimensional dynamical system in less than five milliseconds and solves the optimization problem in under one millisecond, yielding a controller update rate of 100Hz.
{"title":"Safety from Fast, In-the-Loop Reachability with Application to UAVs","authors":"Christian Llanes, Matthew Abate, S. Coogan","doi":"10.1109/iccps54341.2022.00018","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00018","url":null,"abstract":"We present a runtime assurance (RTA) mechanism for ensuring safety of a controlled dynamical system and an application to collision avoidance of two unmanned aerial vehicles (UAVs). We consider a dynamical system controlled by an unverified and potentially unsafe primary controller that might, e.g., lead to collision. The proposed RTA mechanism computes at each time the reachable set of the system under a backup control law. We then develop a novel optimization problem based on control barrier functions that filters the primary controller when necessary in order to keep the system's reachable set within reach of a known, but conservative, safe region. The theory of mixed monotone systems is leveraged for efficient reachable set computation and to achieve a tractable optimization formulation. We demonstrate the proposed RTA mechanism on a dual multirotor UAV case study which requires a fast controller update rate as a result of the small time-scale rotational dynamics. In implementation, the algorithm computes the reachable set of an eight dimensional dynamical system in less than five milliseconds and solves the optimization problem in under one millisecond, yielding a controller update rate of 100Hz.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115490891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00033
Ahmed El Yaacoub, L. Mottola, T. Voigt, Philipp Rümmer
Dynamic software updates enable software evolution and bug fixes to embedded systems without disrupting their run-time operation. Scheduling dynamic updates for safety-critical embedded systems, such as aerial drones, must be done with great care. Otherwise, the system's control loop will be delayed leading to a partial or even complete loss of control, ultimately impacting the dependable operation. We propose an update scheduling algorithm called NeRTA, which schedules updates during the short times when the processor would have been idle. NeRTA consequently avoids the loss of control that would occur if an update delayed the execution of the control loop. The algorithm computes conservative estimations of idle times to determine if an update is possible, but is also sufficiently accurate that the estimated idle time is typically within 15% of the actual idle time.
{"title":"Poster Abstract: Scheduling Dynamic Software Updates in Safety-critical Embedded Systems - the Case of Aerial Drones","authors":"Ahmed El Yaacoub, L. Mottola, T. Voigt, Philipp Rümmer","doi":"10.1109/iccps54341.2022.00033","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00033","url":null,"abstract":"Dynamic software updates enable software evolution and bug fixes to embedded systems without disrupting their run-time operation. Scheduling dynamic updates for safety-critical embedded systems, such as aerial drones, must be done with great care. Otherwise, the system's control loop will be delayed leading to a partial or even complete loss of control, ultimately impacting the dependable operation. We propose an update scheduling algorithm called NeRTA, which schedules updates during the short times when the processor would have been idle. NeRTA consequently avoids the loss of control that would occur if an update delayed the execution of the control loop. The algorithm computes conservative estimations of idle times to determine if an update is possible, but is also sufficiently accurate that the estimated idle time is typically within 15% of the actual idle time.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114244373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00014
Qitong Gao, Stephen L. Schmidt, Karthik Kamaravelu, D. Turner, W. Grill, M. Pajic
Deep brain stimulation (DBS) is an effective procedure to treat motor symptoms caused by nervous system disorders such as Parkinson's disease (PD). Although existing implantable DBS devices can suppress PD symptoms by delivering fixed periodic stimuli to the Basal Ganglia (BG) region of the brain, they are considered inefficient in terms of energy and could cause side-effects. Recently, reinforcement learning (RL)-based DBS controllers have been developed to achieve both stimulation efficacy and energy efficiency, by adapting stimulation parameters (e.g., pattern and frequency of stimulation pulses) to the changes in neuronal activity. However, RL methods usually provide limited safety and performance guarantees, and directly deploying them on patients may be hindered due to clinical regulations. Thus, in this work, we introduce a model-based offline policy evaluation (OPE) methodology to estimate the performance of RL policies using historical data. As a first step, the BG region of the brain is modeled as a Markov decision process (MDP). Then, a deep latent MDP (DL-MDP) model is learned using variational inference and previously collected control trajectories. The performance of RL controllers is then evaluated on the DL-MDP models instead of patients directly, ensuring safety of the evaluation process. Further, we show that our method can be integrated into offline RL frameworks, improving control performance when limited training data are available. We illustrate the use of our methodology on a computational Basal Ganglia model (BGM); we show that it accurately estimates the expected returns of controllers trained following state-of-the-art RL frameworks, outperforming existing OPE methods designed for general applications.
{"title":"Offline Policy Evaluation for Learning-based Deep Brain Stimulation Controllers","authors":"Qitong Gao, Stephen L. Schmidt, Karthik Kamaravelu, D. Turner, W. Grill, M. Pajic","doi":"10.1109/iccps54341.2022.00014","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00014","url":null,"abstract":"Deep brain stimulation (DBS) is an effective procedure to treat motor symptoms caused by nervous system disorders such as Parkinson's disease (PD). Although existing implantable DBS devices can suppress PD symptoms by delivering fixed periodic stimuli to the Basal Ganglia (BG) region of the brain, they are considered inefficient in terms of energy and could cause side-effects. Recently, reinforcement learning (RL)-based DBS controllers have been developed to achieve both stimulation efficacy and energy efficiency, by adapting stimulation parameters (e.g., pattern and frequency of stimulation pulses) to the changes in neuronal activity. However, RL methods usually provide limited safety and performance guarantees, and directly deploying them on patients may be hindered due to clinical regulations. Thus, in this work, we introduce a model-based offline policy evaluation (OPE) methodology to estimate the performance of RL policies using historical data. As a first step, the BG region of the brain is modeled as a Markov decision process (MDP). Then, a deep latent MDP (DL-MDP) model is learned using variational inference and previously collected control trajectories. The performance of RL controllers is then evaluated on the DL-MDP models instead of patients directly, ensuring safety of the evaluation process. Further, we show that our method can be integrated into offline RL frameworks, improving control performance when limited training data are available. We illustrate the use of our methodology on a computational Basal Ganglia model (BGM); we show that it accurately estimates the expected returns of controllers trained following state-of-the-art RL frameworks, outperforming existing OPE methods designed for general applications.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117210351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-05-01DOI: 10.1109/iccps54341.2022.00044
Truls Nyberg, José Manuel Gaspar Sánchez, Christian Pek, Jana Tumova, Martin Törngren
Hidden traffic participants pose a great challenge for autonomous vehicles. Previous methods typically do not use previous obser-vations, leading to over-conservative behavior. In this paper, we present a continuation of our work on reasoning about objects out-side the current sensor view. We aim to demonstrate our recently proposed method on an autonomous platform and evaluate its relia-bility and real-time feasibility when using real sensor data. Showing a significant driving performance increase on a real platform, with-out compromising safety, would be a significant contribution to the field of autonomous driving.
{"title":"Evaluating Sequential Reasoning about Hidden Objects in Traffic","authors":"Truls Nyberg, José Manuel Gaspar Sánchez, Christian Pek, Jana Tumova, Martin Törngren","doi":"10.1109/iccps54341.2022.00044","DOIUrl":"https://doi.org/10.1109/iccps54341.2022.00044","url":null,"abstract":"Hidden traffic participants pose a great challenge for autonomous vehicles. Previous methods typically do not use previous obser-vations, leading to over-conservative behavior. In this paper, we present a continuation of our work on reasoning about objects out-side the current sensor view. We aim to demonstrate our recently proposed method on an autonomous platform and evaluate its relia-bility and real-time feasibility when using real sensor data. Showing a significant driving performance increase on a real platform, with-out compromising safety, would be a significant contribution to the field of autonomous driving.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125866207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-28DOI: 10.48550/arXiv.2203.15127
Michael Wilbur, S. U. Kadir, Youngseo Kim, Geoffrey Pettet, Ayan Mukhopadhyay, Philip Pugliese, S. Samaranayake, Aron Laszka, Abhishek Dubey
Many transit agencies operating paratransit and microtransit ser-vices have to respond to trip requests that arrive in real-time, which entails solving hard combinatorial and sequential decision-making problems under uncertainty. To avoid decisions that lead to signifi-cant inefficiency in the long term, vehicles should be allocated to requests by optimizing a non-myopic utility function or by batching requests together and optimizing a myopic utility function. While the former approach is typically offline, the latter can be performed online. We point out two major issues with such approaches when applied to paratransit services in practice. First, it is difficult to batch paratransit requests together as they are temporally sparse. Second, the environment in which transit agencies operate changes dynamically (e.g., traffic conditions can change over time), causing the estimates that are learned offline to become stale. To address these challenges, we propose a fully online approach to solve the dynamic vehicle routing problem (DVRP) with time windows and stochastic trip requests that is robust to changing environmental dynamics by construction. We focus on scenarios where requests are relatively sparse-our problem is motivated by applications to paratransit services. We formulate DVRP as a Markov decision process and use Monte Carlo tree search to evaluate actions for any given state. Accounting for stochastic requests while optimizing a non-myopic utility function is computationally challenging; indeed, the action space for such a problem is intractably large in practice. To tackle the large action space, we leverage the structure of the problem to design heuristics that can sample promising actions for the tree search. Our experiments using real-world data from our partner agency show that the proposed approach outperforms existing state-of-the-art approaches both in terms of performance and robustness.
{"title":"An Online Approach to Solve the Dynamic Vehicle Routing Problem with Stochastic Trip Requests for Paratransit Services","authors":"Michael Wilbur, S. U. Kadir, Youngseo Kim, Geoffrey Pettet, Ayan Mukhopadhyay, Philip Pugliese, S. Samaranayake, Aron Laszka, Abhishek Dubey","doi":"10.48550/arXiv.2203.15127","DOIUrl":"https://doi.org/10.48550/arXiv.2203.15127","url":null,"abstract":"Many transit agencies operating paratransit and microtransit ser-vices have to respond to trip requests that arrive in real-time, which entails solving hard combinatorial and sequential decision-making problems under uncertainty. To avoid decisions that lead to signifi-cant inefficiency in the long term, vehicles should be allocated to requests by optimizing a non-myopic utility function or by batching requests together and optimizing a myopic utility function. While the former approach is typically offline, the latter can be performed online. We point out two major issues with such approaches when applied to paratransit services in practice. First, it is difficult to batch paratransit requests together as they are temporally sparse. Second, the environment in which transit agencies operate changes dynamically (e.g., traffic conditions can change over time), causing the estimates that are learned offline to become stale. To address these challenges, we propose a fully online approach to solve the dynamic vehicle routing problem (DVRP) with time windows and stochastic trip requests that is robust to changing environmental dynamics by construction. We focus on scenarios where requests are relatively sparse-our problem is motivated by applications to paratransit services. We formulate DVRP as a Markov decision process and use Monte Carlo tree search to evaluate actions for any given state. Accounting for stochastic requests while optimizing a non-myopic utility function is computationally challenging; indeed, the action space for such a problem is intractably large in practice. To tackle the large action space, we leverage the structure of the problem to design heuristics that can sample promising actions for the tree search. Our experiments using real-world data from our partner agency show that the proposed approach outperforms existing state-of-the-art approaches both in terms of performance and robustness.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129170096","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-11DOI: 10.48550/arXiv.2203.06220
Jihoon Yun, Sangeeta Srivastava, Dhrubojyoti Roy, Nathan Stohs, C. Mydlarz, Mahiny A. Salman, Bea Steers, J. Bello, Anish Arora
The Sounds of New York City (SONYC) wireless sensor network (WSN) has been fielded in Manhattan and Brooklyn over the past five years, as part of a larger human-in-the-loop cyber-physical control system for monitoring, analyzing, and mitigating urban noise pollution. We describe the evolution of the 2-tier SONYC WSN from an acoustic data collection fabric into a 3-tier in situ noise complaint monitoring WSN, and its current evaluation. The added tier consists of long range (LoRa), multi-hop networks of a new low-power acoustic mote, MKII (“Mach 2”), that we have designed and fabricated. MKII motes are notable in three ways: First, they advance machine learning capability at mote-scale in this application domain by introducing a real-time Convolutional Neural Network (CNN) based embedding model that is competitive with alternatives while also requiring 10x lesser training data and ~2 orders of magnitude fewer runtime resources. Second, they are conveniently deployed relatively far from higher-tier base station nodes without assuming power or network infrastructure support at operationally relevant sites (such as construction zones), yielding a relatively low-cost solution. And third, their networking is frequency agile, unlike conventional LoRa networks: it tolerates in a distributed, self-stabilizing way the variable external interfer-ence and link fading in the cluttered 902-928MHz ISM band urban environment by dynamically choosing good frequencies using an efficient new method that combines passive and active measure-ments.
{"title":"Infrastructure-free, Deep Learned Urban Noise Monitoring at ~100mW","authors":"Jihoon Yun, Sangeeta Srivastava, Dhrubojyoti Roy, Nathan Stohs, C. Mydlarz, Mahiny A. Salman, Bea Steers, J. Bello, Anish Arora","doi":"10.48550/arXiv.2203.06220","DOIUrl":"https://doi.org/10.48550/arXiv.2203.06220","url":null,"abstract":"The Sounds of New York City (SONYC) wireless sensor network (WSN) has been fielded in Manhattan and Brooklyn over the past five years, as part of a larger human-in-the-loop cyber-physical control system for monitoring, analyzing, and mitigating urban noise pollution. We describe the evolution of the 2-tier SONYC WSN from an acoustic data collection fabric into a 3-tier in situ noise complaint monitoring WSN, and its current evaluation. The added tier consists of long range (LoRa), multi-hop networks of a new low-power acoustic mote, MKII (“Mach 2”), that we have designed and fabricated. MKII motes are notable in three ways: First, they advance machine learning capability at mote-scale in this application domain by introducing a real-time Convolutional Neural Network (CNN) based embedding model that is competitive with alternatives while also requiring 10x lesser training data and ~2 orders of magnitude fewer runtime resources. Second, they are conveniently deployed relatively far from higher-tier base station nodes without assuming power or network infrastructure support at operationally relevant sites (such as construction zones), yielding a relatively low-cost solution. And third, their networking is frequency agile, unlike conventional LoRa networks: it tolerates in a distributed, self-stabilizing way the variable external interfer-ence and link fading in the cluttered 902-928MHz ISM band urban environment by dynamically choosing good frequencies using an efficient new method that combines passive and active measure-ments.","PeriodicalId":340078,"journal":{"name":"2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117283742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：