Predicting Away Robot Control Latency

Behnke, Sven; Egorova, Anna; Gloye, Alexander; Rojas, Raúl; Simon, Mark

doi:10.1007/978-3-540-25940-4_70

Cited by 37 publications

(17 citation statements)

References 8 publications

(8 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This world model allows for extrapolating the movement of all relevant objects into the future . MLPs for predicting parts of the world model have been included as well, following an idea of (Behnke et al 2003). Thus, decisionmaking in our architecture never works on the state of the environment as it was sensed by the sensors, but always uses the state extrapolated to the moment in future when a decision actually will become active.…”

Section: An Architecture For Soccer-playing Robotsmentioning

confidence: 99%

Reinforcement learning for robot soccer

et al. 2009

View full text Add to dashboard Cite

Batch reinforcement learning methods provide a powerful framework for learning efficiently and effectively in autonomous robots. The paper reviews some recent work of the authors aiming at the successful application of reinforcement learning in a challenging and complex domain. It discusses several variants of the general batch learning framework, particularly tailored to the use of multilayer perceptrons to approximate value functions over continuous state spaces. The batch learning framework is successfully used to learn crucial skills in our soccer-playing robots participating in the RoboCup competitions. This is demonstrated on three different case studies.

show abstract

Section: An Architecture For Soccer-playing Robotsmentioning

confidence: 99%

Reinforcement learning for robot soccer

et al. 2009

View full text Add to dashboard Cite

show abstract

“…MLPs have been shown to have good mapping capabilities for modeling both real and virtual entities in a number of application domains, including distributed military simulations [Henninger et al 2001], networked virtual environments [Sas et al 2003], networked multiplayer computer games [Thurau et al 2003], and autonomous robot soccer [Behnke et al 2003]. In our approach, each successive MLP accepts the same input column vector as its predecessors, but is trained to predict over a longer prediction horizon independently of the outputs from other networks.…”

Section: Neural-network Predictorsmentioning

confidence: 99%

“…Thus, for the purposes of training the set of neural network predictors, entity velocity and orientation are modeled as 2-dimensional state vectors (i.e., d = 2). As noted by Behnke et al [2003], we encode the entity orientation vector as a point (i.e., a sine and cosine coordinate pair) on a unit circle centered at the entity's current position to avoid the training complications associated with the discontinuity between π and -π when encoding a single angle. All measurements are defined in terms of torque world units (TWUs).…”

Section: Overviewmentioning

confidence: 99%

Multistep-ahead neural-network predictors for network traffic reduction in distributed interactive applications

McCoy

Ward

McLoone

et al. 2007

ACM Trans. Model. Comput. Simul.

View full text Add to dashboard Cite

Predictive contract mechanisms such as dead reckoning are widely employed to support scalable remote entity modeling in distributed interactive applications (DIAs). By employing a form of controlled inconsistency, a reduction in network traffic is achieved. However, by relying on the distribution of instantaneous derivative information, dead reckoning trades remote extrapolation accuracy for low computational complexity and ease-of-implementation. In this article, we present a novel extension of dead reckoning, termed neuro-reckoning, that seeks to replace the use of instantaneous velocity information with predictive velocity information in order to improve the accuracy of entity position extrapolation at remote hosts. Under our proposed neuro-reckoning approach, each controlling host employs a bank of neural network predictors trained to estimate future changes in entity velocity up to and including some maximum prediction horizon. The effect of each estimated change in velocity on the current entity position is simulated to produce an estimate for the likely position of the entity over some short time-span. Upon detecting an error threshold violation, the controlling host transmits a predictive velocity vector that extrapolates through the estimated position, as opposed to transmitting the instantaneous velocity vector. Such an approach succeeds in reducing the spatial error associated with remote extrapolation of entity state. Consequently, a further reduction in network traffic can be achieved. Simulation results conducted using several human users in a highly interactive DIA indicate significant potential for improved scalability when compared to the use of IEEE DIS standard dead reckoning. Our proposed neuro-reckoning framework exhibits low computational resource overhead for real-time use and can be seamlessly integrated into many existing dead reckoning mechanisms. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or direct commercial advantage and that copies show this notice on the first page or initial screen of a display along with the full citation. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, to redistribute to lists, or to use any component of this work in other works requires prior specific permission and/or a fee. Permissions may be requested from Publications Dept., ACM, Inc., 2 Penn Plaza, Suite 701, New York, NY 10121-0701 USA, fax +1 (212)

show abstract

“…A lot of work is done to deal with this delay with observers and predictors [10][11][12][13][14] and to prove the stability of systems suffering from variable transmission delays [15,16]. There is not much literature about wireless control applications.…”

Section: Introductionmentioning

confidence: 99%

Low‐Latency Wireless Data Transfer for Motion Control

Boeij

Haazen

Smulders

et al. 2009

Journal of Control Science and Engineering

View full text Add to dashboard Cite

This paper discusses a new approach for wireless motion control. Existing wireless techniques suffer from large closed loop delays of several milliseconds, which is unacceptable in precision motion systems. These large delays are mainly caused by the protocol used, since these are optimized for transferring large amounts of data, not to minimize transmission delay. A new protocol and wireless system are proposed that reduce the closed loop transmission delay to less than 300 microseconds. The system has been verified in a real control system, and measurements show the performance, which is more than ten times better than existing techniques.

show abstract

Predicting Away Robot Control Latency

Cited by 37 publications

References 8 publications

Reinforcement learning for robot soccer

Reinforcement learning for robot soccer

Multistep-ahead neural-network predictors for network traffic reduction in distributed interactive applications

Low‐Latency Wireless Data Transfer for Motion Control

Contact Info

Product

Resources

About