8,569 Hits in 3.8 sec

Model-free robot anomaly detection

Rachel Hornung, Holger Urbanek, Julian Klodmann, Christian Osendorfer, Patrick van der Smagt
2014 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems  
Safety is one of the key issues in the use of robots, especially when human-robot interaction is targeted.  ...  No sufficiently fast and reliable methods exist which can early detect faults in the abundance of sensor and controller data.  ...  However, this technique cannot detect faults exhibiting a similar intra-data relationship as the training data. In 2011 [14] have introduced another model-free approach to anomaly detection.  ... 
doi:10.1109/iros.2014.6943078 dblp:conf/iros/HornungUKOS14 fatcat:6juxqsf5yvb4vez2lxzcwbhxoq

Secure Planning Against Stealthy Attacks via Model-Free Reinforcement Learning [article]

Alper Kamil Bozkurt, Yu Wang, Miroslav Pajic
2021 arXiv   pre-print
We model the attacker as an agent who has the full knowledge of the controller as well as the employed intrusion-detection system and who wants to prevent the controller from performing tasks while staying  ...  We then show that the planning problem, described formally as the problem of satisfying an LTL formula in a stochastic game, can be solved via model-free reinforcement learning when the environment is  ...  Furthermore, to allow the use of our security-aware planning methodology for robotic systems with unknown models, we adopt a model-free RL approach to solve the game between the attacker and controller  ... 
arXiv:2011.01882v2 fatcat:7mue56sg3bfhha45udqkbigkqq

An efficient statistical strategy to monitor a robot swarm

Fouzi Harrou, Belkacem Khaldi, Ying Sun, Foudil Cherif
2019 IEEE Sensors Journal  
The principal component analysis model is employed to generate residuals for anomaly detection.  ...  This letter deals with the problem of detecting anomalies in a robot swarm.  ...  Based on fault-free measurements, the reference PCA model is designed and then adopted for fault detection.  ... 
doi:10.1109/jsen.2019.2950695 fatcat:maewzbkkmnhs5l65oncljybfwm

Variational Inference for On-line Anomaly Detection in High-Dimensional Time Series [article]

Maximilian Soelch, Justin Bayer, Marvin Ludersdorfer, Patrick van der Smagt
2016 arXiv   pre-print
We apply a Stochastic Recurrent Network (STORN) to learn robot time series data. Our evaluation demonstrates that we can robustly detect anomalies both off- and on-line.  ...  Approximate variational inference has shown to be a powerful tool for modeling unknown complex probability distributions.  ...  We have shown that the new approach is able to detect anomalies in robot time series data with remarkably high precision.  ... 
arXiv:1602.07109v5 fatcat:nmmg2qysirc5tobstgtmhzg7pm

Statistical detection of faults in swarm robots under noisy conditions

Fouzi Harrou, Belkacem Khaldi, Ying Sun, Foudil Cherif
2018 2018 6th International Conference on Control Engineering & Information Technology (CEIT)  
Fault detection plays an important role in supervising the operation of robotic swarm systems. If faults are not detected, they can considerably affect the performance of the robot swarm.  ...  component analysis model.  ...  Once the model is developed from the training data, it would be used with the proposed WM-EWMA approach to detect anomalies in the testing data.  ... 
doi:10.1109/ceit.2018.8751862 fatcat:xwnhbq5x45cm7jwmw7usonqvwy

Anomaly Detection in Smart Manufacturing with an Application Focus on Robotic Finishing Systems: A Review [article]

Tareq Tayeh, Abdallah Shami
2021 arXiv   pre-print
In this paper, an overview of the components, benefits, challenges, methods, and open problems of anomaly detection in smart manufacturing and robotic finishing systems are discussed.  ...  However, with the deployment of anomaly detection systems, there are many aspects to be considered.  ...  observation only, where each data point is analyzed exactly once by the anomaly detection model.  ... 
arXiv:2107.05053v2 fatcat:ct7hofnyr5e33f4g5sl5mvjt6q

Plan execution monitoring through detection of unmet expectations about action outcomes

Juan Pablo Mendoza, Manuela Veloso, Reid Simmons
2015 2015 IEEE International Conference on Robotics and Automation (ICRA)  
Modeling the effects of actions based on the state of the world enables robots to make intelligent decisions in different situations. However, it is often infeasible to have globally accurate models.  ...  Furthermore, expectations about the world are often stochastic in robotics, making the discovery of model-world discrepancies non-trivial.  ...  Model-free RL approaches, such as Q-Learning [6] and policy gradient descent [7] , are capable of improving robot performance without explicitly modeling the world.  ... 
doi:10.1109/icra.2015.7139646 dblp:conf/icra/MendozaVS15 fatcat:7qcuqgzfwjalfpdwn5vneo2oke

Predictive Maintenance: An Autoencoder Anomaly-Based Approach for a 3 DoF Delta Robot

Kiavash Fathi, Hans Wernher van de Venn, Marcel Honegger
2021 Sensors  
In the proposed method, autoencoders (AEs) are used to predict when maintenance is required based on the signal sequence distribution and anomaly detection, which is vital when no R2F data are available  ...  In this paper, a 3 DoF delta robot used for pick and place task is studied.  ...  In addition, the proposed architecture is superior to analytical models given the following reasons. Using AEs, a movement dependent anomaly detection model can be trained.  ... 
doi:10.3390/s21216979 pmid:34770289 pmcid:PMC8588519 fatcat:sdjrgnizjzgxhhpwflqrsmotma

Improving robot manipulation with data-driven object-centric models of everyday forces

Advait Jain, Charles C. Kemp
2013 Autonomous Robots  
We also demonstrate that two distinct robots can use forces captured from people opening doors to better detect anomalous forces.  ...  cabinet), and haptically detect anomalous forces while opening a door, even when opening a specific door for the first time.  ...  Haptic recognition and anomaly detection In contrast to our use of data-driven object-centric models, previous research on haptic recognition and anomaly detection for robot manipulation has often used  ... 
doi:10.1007/s10514-013-9344-1 fatcat:63nzqjeq45flvfz3vwopaowjc4

Multi-Modal Anomaly Detection for Unstructured and Uncertain Environments [article]

Tianchen Ji, Sri Theja Vuppala, Girish Chowdhary, Katherine Driggs-Campbell
2020 arXiv   pre-print
To achieve high-levels of autonomy, modern robots require the ability to detect and recover from anomalies and failures with minimal human supervision.  ...  Multi-modal sensor signals could provide more information for such anomaly detection tasks; however, the fusion of high-dimensional and heterogeneous sensor modalities remains a challenging problem.  ...  We would like to thank EarthSense for providing the field robot data, and the reviewers for their thorough and constructive comments.  ... 
arXiv:2012.08637v1 fatcat:yioxnxwktfdmtpceuk5zsdwicq

Learning Skills to Patch Plans Based on Inaccurate Models [article]

Alex LaGrassa, Steven Lee, Oliver Kroemer
2020 arXiv   pre-print
First, we use a sub-optimal model-based planner to perform a task until model failure is detected.  ...  Next, we learn a local model-free policy from expert demonstrations to complete the task in regions where the model failed.  ...  ACKNOWLEDGEMENTS We thank Kevin Zhang, Jacky Liang, Mohit Sharma, and many others for providing the infrastructure needed for the robot experiments.  ... 
arXiv:2009.13732v1 fatcat:5xe67rfvkzbm7hcmi7vvchcaki

Exploring Granger causality in dynamical systems modeling and performance monitoring

Homagni Saha, Chao Liu, Zhanhong Jiang, Soumik Sarkar
2018 2018 IEEE Conference on Decision and Control (CDC)  
We apply this new framework for anomaly detection and root cause analysis in a robotic platform. CHAPTER 1.  ...  In (52), the authors propose to generate adaptive thresholds using locally linear models (LLM) and Model Error Modeling (MEM) techniques for fault detection on two wheeled mobile robot.  ... 
doi:10.1109/cdc.2018.8619530 dblp:conf/cdc/SahaLJS18 fatcat:c334vdi535ec5f6cab2zrt3isu

Online data-driven anomaly detection in autonomous robots

Eliahu Khalastchi, Meir Kalech, Gal A. Kaminka, Raz Lin
2014 Knowledge and Information Systems  
ODDAD is suitable for the dynamic nature of autonomous robots since it declares a fault based only on data collected online. In addition, it is unsupervised, model free and domain independent.  ...  Not all faults can be known in advance, and hence, anomaly detection is required. In this paper, we present an online datadriven anomaly detection approach (ODDAD) for autonomous robots.  ...  Another model-based approach for anomaly detection is model-based reasoning (e.g., [10, 28] ).  ... 
doi:10.1007/s10115-014-0754-y fatcat:jnjwxbgwo5duxky2ni7y2lnazy

Auto-Encoding Robot State Against Sensor Spoofing Attacks

Sean Rivera, Sofiane Lagraa, Antonio Ken Iannillo, Radu State
2019 2019 IEEE International Symposium on Software Reliability Engineering Workshops (ISSREW)  
In this paper, we propose a novel anomaly detection approach for sensor spoofing attacks, based on an auto-encoder architecture.  ...  We tested our anomaly detection approach against several types of spoofing attacks comparing four different compression rates for the auto-encoder.  ...  Learning Phase In the initial phase, we train the models over a attack-free dataset, to determine what is normal LiDAR data.  ... 
doi:10.1109/issrew.2019.00080 dblp:conf/issre/RiveraLIS19 fatcat:iizaxhjeyjavli7jbxdnptosf4

Detecting Bot Activity in the Ethereum Blockchain Network [article]

Morit Zwang, Shahar Somin, Alex 'Sandy' Pentland, Yaniv Altshuler
2018 arXiv   pre-print
This attribute makes the Ethereum network a breeding space for activity by software robots (bots).  ...  In this work we demonstrate how bot detection can be implemented using a network theory approach.  ...  In some cases, anomalies from the power law model in human interaction networks may be evidence for emergency events [10] .  ... 
arXiv:1810.01591v1 fatcat:uyzodwiqvbb3zblaepdjl2dilm
« Previous Showing results 1 — 15 out of 8,569 results