Professor Sandor M. Veres

Director of the Autonomous Systems and Robotics Research Group

Autonomous Systems and Robotics Research Group

SVeresAddress:
Professor Sandor M. Veres, MEng, PhD, CEng, CMathSci, MIEEE, MIET
Department of Automatic Control and Systems Engineering
University of Sheffield
Sheffield
S1 3JD
Tel: (+44) (0)114 222 5652
Fax: (+44) (0)114 222 5683
Email: s.veres@sheffield.ac.uk
Room: C07a, Amy Johnson Building


Samples of My Simulated and Laboratory Based Autonomous Systems Videos

Research interests
Grants
Professional activities and recognition
Key publications since 2010

Research interests

Optimal decision making in autonomous systems, mobile robotics, agent supervised feedback control systems, architectures and programming of reasoning based intelligent agents, formation flying control of robotic vehicles, formal modelling and verification of systems by model checking, fault tolerant control systems, automated processes of data based modelling, robust adaptive control, controller tuning and system identification, satellite dynamics and control, active control of sound and vibration control.

Autonomous Systems and Robotics Research Group

Formerly creator and coordinator of the Southampton based Centre for Complex Autonomous Systems Engineering.

Back to top

Grants


Current EPSRC Support
 
No. Title EPSRC Contribution
EP/J011843/2 Reconfigurable Autonomy (P Sheffield) 1.31m (Liverpool-Surrey joint project)
EP/J011894/2 Distributed Sensing, Control and Decision Making in Multiagent autonomous Systems (P) 1.23m (joint project with Southampton)
EP/L024942/1 Verifiable Autonomy (P) 1.41m (joint project with Liverpool and Bristol)
 
Previous EPSRC Support
 
No. Title EPSRC Contribution
EP/F037570/1 Engineering Autonomous Space Software (P) 421k
EP/E02677X/1 Methods of Reliability Control for Autonomous Underwater Vehicles (P) 307k
EP/C517458/1 Equipment For Satellite Formation Flying Control System Verification (P) 80k
GR/S25081/01 Constrained Control Methods for Autonomous Formation Flying Systems (P) 158k
GR/N32297/01 Robust Frequency Selective Adaptive Control of Enclosure Vibration (P) 179k
GR/J10846/01 Robust Adaptive Control by Parameter Bounding Techniques (P) 53k
GR/H77200/01 Performance Analysis and Algrorithm Development for Parameter and State Bounding (C) 100k

Key: (P)=Principal Investigator, (C)=Co-Investigator

Back to top

Professional activities and recognition

  • Member of the EPSRC College
  • Subject Editor of the Journal of Adaptive Control and Signal Processing
  • Editorial Board member of the IMechE Journal of Systems and Control Engineering
  • Member of the Advisory Board of Unmanned Systems journal.
  • Referee for many journals
  • Member of International Program Committees of many conferences
  • Advisor and consultant to companies both in the UK and internationally

Back to top

Books

Synergy and Duality of Identification and Control thumbnail Sandor M Veres and Derek S Wall:
Synergy and Duality of Identification and Control

Abstract:
Synergy and Duality of Identification and Control examines the relationship between  modelling and control of dynamic systems. Both stochastic and worst-case design  approaches are presented. System identification is introducted in worst-case and stochastic frameworks. Interaction between modelling and control is studied at four levels: separate identification and control design, iterative identification and control design, dual control and synergistic interaction of modelling and control. The last part of the book focuses on fast learning control systems via parametric models. This comprehensive monograph is equally suitable for graduate students as well as for  researchers in the fields of control engineering and digital signal processing.

Hardcover: 528 pages
Publisher: Taylor & Francis Ltd (28 Oct 1999)
Language: English
ISBN-10: 0748407731
Structure Selection of Stochastic Dynamic Systems: An Information Criterion Approach thumbnail Sandor M Veres:
Structure Selection of Stochastic Dynamic Systems: An Information Criterion Approach

Abstract:
This book gives a reliable review on structure selection of stochastic dynamic systems  using information criteria AIC, BIC, o and stochastic complexity. After theoretical investigations many simulations are estimators, which illustrate both the effectiveness
and the limitations of these methods. The reader can gain his or her own experience on the"working" of many methods (associated with different parameter estimators) using
the demonstration disk which can be run on most IBM-compatible personal computers. The book will be helpful to anybody interested in applying automated methods of model-structure selection inn control engineering, in time series analysis or in signal processing.

Hardcover: 356 pages
Publisher: CRC Press (1 Jan 1991)
Language: English
ISBN-10: 2881247156
Active Sound and Vibration Control thumbnail M Osman Tokhi and Sandor M Veres (Editors)
Active Sound and Vibration Control

Abstract:
This book presents the established fundamentals in the area of active sound and vibration  control as well as exploring the new and emerging technologies and techniques. There has  been a considerable amount of effort devoted to the development and realisation of  methodologies for control of sound and vibration, and this book covers the latest theoretical, algorithmic and practical applications including: noise control in 3D propagation, adaptive algorithms, prediction, processing and tuning, neuro-active control,  control of microvibrations, and noise reduction in locomotives and vehicles.

Hardcover: 450 pages
Publisher: Institution of Engineering and Technology; Illustrated. edition (25 Mar 2002)
Language: English
ISBN-10: 0852960387
Natural Language Programming of Agents and Robotic Devices thumbnail Sandor M Veres
Natural Language Programming of Agents and Robotic Devices

Abstract:
This book provides examples of how conceptual structures can be built up for the purpose of developing shared understanding between man and machine. The medium of this understanding is a version of English that is carefully defined and tailored for a particular application. The name of this type of English is called system English, sEnglish for short. The book outlines the principles and its use with electronic personal assistants and intelligent software agents.

Paperback: 184 pages
Publisher: SysBrain Ltd; 1st edition (6 Jun 2008)
Language: English
ISBN-10: 0955841704

Back to top

Key research topics and publications since 2010

A central tool of my research is natural language programming of control procedures and modelling descriptions for intelligent control agents in sEnglish® that is currently being used by UK based partner companies of the Autonomous Intelligent Systems EPSRC Programme, by Thales and by companies worldwide (www.senglish.net www.sysbrain.org/wiki, Media Reports). sEnglish® is not a main research topic but is used as a tool to support and to help create machine-understandable research publications. My areas of interest are:

  • Reader-executable and machine readable technical papers for journals and the world wide web
  • Navigation and world modelling/SLAM by autonomous vehicles and robots for situational awarness
  • Formal verification and functional safety of autononous control systems and robots, the artificial moral agent
  • Learning and adaptive control methods as part of the knowledge base of intelligent machines
  • Reconfigurable, self-organising-agent based autonomous systems formats and standards
  • Distributed sensing, control and decision making in autonomous control systems
  • Novel deliberative agent architectures of the belief-desire-intention (BDI) type for autonomous systems and robots

Started work at the Department of Automatic Control and Systems Engineering, University of Seffield, in November 2012, prior to that created the Autonomous Vehicle Systems Lab and established the Centre for Complex Autonomous Systems Engineering at Southampton University.

  Details of Publication Abstract
2016 A stochastically verifiable autonomous control architecture with reasoning (2016), by P Izzo, H Qu and S M Veres, Proc. IEEE Conf. Decision and Control, Dec. 2016, Las Vegas.
http://arxiv.org/abs/1611.03372
A new agent architecture called Limited Instruction Set Agent (LISA) is introduced for autonomous control. The new architecture is based on previous implementations of AgentSpeak and it is structurally simpler than its predecessors with the aim of facilitating design-time and run-time verification methods. The process of abstracting the LISA system to two different types of discrete probabilistic models (DTMC and MDP) is investigated and illustrated. The LISA system provides a tool for complete modelling of the agent and the environment for probabilistic verification. The agent program can be automatically compiled into a TMC or a MDP model for verification with Prism. The automatically generated Prism model can be used for both design-time and run-time verification. The run-time verification is investigated and illustrated in the LISA system as an internal modelling mechanism for prediction of future outcomes.
2016 Verification of logical consistency in robotic reasoning (2016), by H Qu and S M Veres. Robotics and Autonomous Systems, Vol. 83(2016), 44-56.
http://arxiv.org/abs/1611.03322
Most autonomous robotic agents use logic inference to keep themselves to safe and permitted behaviour. Given a set of rules, it is important that the robot is able to establish the consistency between its rules, its perception-based beliefs, its planned actions and their consequences. This paper investigates how a robotic agent can use model checking to examine the consistency of its rules, beliefs and actions. A rule set is modelled by a Boolean evolution system with synchronous semantics, which can be translated into a labelled transition system (LTS). It is proven that stability and consistency can be formulated as computation tree logic (CTL) and linear temporal logic (LTL) properties. Two new algorithms are presented to perform realtime consistency and stability checks respectively. Their implementation provides us a computational tool, which can form the basis of efficient consistency checks on-board robots.
2016 Fictitious play for cooperative action selection in robot teams. (2016), by M Smyrnakis and S M Veres, IFAC Journal of Engineering Applications of Artificial Intelligence. Vol. 56, p. 14-29. A game theoretic distributed decision making approach is presented for the problem of control e↵ort allocation in a robotic team based on a novel variant of fictitious play. The proposed learning process allows the robots to accomplish their objectives by coordinating their actions in order to e ciently complete their tasks. In particular, each robot of the team predicts the other robots’ planned actions while making decisions to maximise their own expected reward that depends on the reward for joint successful completion of the task. Action selection is interpreted as an n-player cooperative game. The approach presented can be seen as part of the Belief Desire Intention (BDI) framework, also can address the problem of cooperative, legal, safe, considerate and emphatic decisions by robots if their individual and group rewards are suitably defined. After theoretical analysis the performance of the proposed algorithm is tested on four simulation scenarios. The first one is a coordination game between two material handling robots, the second one is a warehouse patrolling task by a team of robots, the third one presents a coordination mechanism between two robots that carry a heavy object on a corridor and the fourth one is an example of coordination on a sensors network.
2016 Improved system identification using artificial neural networks and analysis of individual differences in responses of an identified neuron (2016), by A. Costalago Meruelo, D M Simpson and P L Newland, J. Neural Networks, Vol. 75, pp 56-65. Mathematical modelling is used routinely to understand the coding properties and dynamics of responses of neurons and neural networks. Here we analyse the effectiveness of Artificial Neural Networks (ANNs) as a modelling tool for motor neuron responses. We used ANNs to model the synaptic responses of an identified motor neuron, the fast extensor motor neuron, of the desert locust in response to displacement of a sensory organ, the femoral chordotonal organ, which monitors movements of the tibia relative to the femur of the leg. The aim of the study was threefold: first to determine the potential value of ANNs as tools to model and investigate neural networks, second to understand the generalisation properties of ANNs across individuals and to different input signals and third, to understand individual differences in responses of an identified neuron. A metaheuristic algorithm was developed to design the ANN architectures. The performance of the models generated by the ANNs was compared with those generated through previous mathematical models of the same neuron. The results suggest that ANNs are significantly better than LNL and Wiener models in predicting specific neural responses to Gaussian White Noise, but not significantly different when tested with sinusoidal inputs. They are also able to predict responses of the same neuron in different individuals irrespective of which animal was used to develop the model, although notable differences between some individuals were evident.
2016 Practical Verification of Decision-Making in Agent-Based Autonomous Systems. (2016), by LA Dennis, M Fisher, N Lincoln, A Lisitsa, S M Veres . J. Automated Software Engineering (2016) Vol 23:305–359. We present a verification methodology for analysing the decision-making component in agent-based hybrid systems. Traditionally hybrid automata have been used to both implement and verify such systems, but hybrid automata based modelling, programming and verification techniques scale poorly as the complexity of discrete decision-making increases making them unattractive in situations where complex logical reasoning is required. In the programming of complex systems it has, therefore, become common to separate out logical decision-making into a separate, discrete, component. However, verification techniques have failed to keep pace with this development.We are exploring agent-based logical components and have developed a model checking technique for such components which can then be composed with a separate analysis of the continuous part of the hybrid system. Among other things this allows program model checkers to be used to verify the actual implementation of the decision-making in hybrid autonomous systems.
2016 Verification-driven design and programming of autonomous robots (2016), by P Izzo , H Qu, O McAree, S M Veres. Robotics and Autonomous Systems (submitted) The limited instruction set agent (LISA), belonging to the AgentSpeak/Jason belief-desire-intention (BDI) family, is analysed for its verifiability as it operates in a physical/social environment. A limited instruction set is used to facilitate design-time and run-time verification methods while retaining the advantageous features of the BDI approach to programming. The results show that LISA programs can be compiled into a discrete time Markov chain or a Markov decision process to model agent interactions with the environment. This enables the use of available probabilistic verification tools such as Prism for verification of robotic agents. The presented theory of the LISA programming system provides a complete solution-set for modelling the agent’s decision making and the environment. The abstractions of sensing, perception, decision making and feedback controllers is facilitated by Natural Language Programming in sEnglish. The automatically compiled models in Prism can be used for both design-time and run-time verification by the agent. Runtime verification is investigated and illustrated as a realtime modelling mechanism in Prism, automatically derived from relevant parts of the LISA program, for probabilistic prediction and value assessment of future outcomes based on alternatives for future actions.
2016 An Intelligent Agent Supervised Reconfigurable Autopilot System (2016), by S. Tantrairatn and S M Veres, AIAA Journal of Guidance, Control, and Dynamics (submitted). A self reconfiguring autopilot system is presented which is based on a rational agent framework that integrates decision making with abstractions of sensing and actions for next generation unmanned aerial vehicles. The objective of the new intelligent control system is to provide advanced capabilities of self-tuning and adaptation in the presence of failures and adverse flight conditions. This system is achieved through onboard dynamical monitoring and estimation associated with controller switching and tuning by the agent. The agent can handle an untuned autopilot or retune the autopilot when dynamical changes occur due to aerodynamic and onboard system changes. The system integrates dynamical modelling, hybrid adaptive control, model validation, ight condition diagnosis, control performance evaluation through software agent development. An important feature of the agent is its abstractions from real-time measurements and also its abstractions about a simulated future which is hypothetical on its control actions. The methods are demonstrated in simulation on a benchmark performance of an Aerosonde UAV (MATLAB/Simulink).
2016 Consensus in cooperative behavior of robot populations (2016), by M Smyrnakis, D Bauso, PA Trodden and S M Veres . IEEE Trans. Automatic Control (submitted). Convergence and equilibrium properties of game theoretic learning algorithms is proven in robot populations while using reward/cost models of cooperation among robotic agents. The contribution of this paper is three-fold. First, new models for robot cooperation are proposed by combining regret based learning methods, network evolution models and various consensus protocols. Second, results from mean-field game theory are used in order to show the asymptotic second moment boundedness in the variation of the cooperative behaviour. Third, the conditions under which scale-free networks can emerge from the proposed algorithms is studied in simulations. In addition the behaviour of the proposed models are studied in networks with a fixed number of nodes and a single lane traffic flow case study.
2016 SMCL – Stochastic Model Checker for Learning in Games (2016), by H Qu, M. Smyrnakis and S M Veres, IEEE Trans. on Cybernetics (submitted) A stochastic model checker is presented for analysing the performance of game-theoretic learning algorithms. The method enables the comparison of short-term behaviour of learning algorithms intended for practical use. The procedure of comparison is automated and it can be tuned for accuracy and speed. Users can choose from among various learning algorithms to select a suitable one for a given practical problem. The powerful performance of the method is enabled by a novel behaviour-similarity-relation, which compacts large state spaces into small ones. The stochastic model checking tool is tested on a set of examples classified into four categories to demonstrate the effectiveness of selecting suitable algorithms for distributed decision making.
2016 Learning of cooperative behavior in robot populations (2016), by M. Smyrnakis, D Bauso, P. Trodden and S. M. Veres. Proc. European Control Conference, Denmark. This paper addresses convergence and equilibrium properties of game theoretic learning algorithms in robot populations using simple and broadly applicable reward/cost models of cooperation between robotic agents. New models for robot cooperation are proposed by combining regret based learning methods and network evolution models. Results of mean-field game theory are employed in order to show the asymptotic second moment boundedness in the variation of cooperative behaviour. The behaviour of the proposed models are tested in simulation results, which are based on sample networks and a single lane traffic flow case study.
2016 Lateral Control of vehicle platoons with on-board sensing and inter-vehicle communication, (2016), by O McAree and S M Veres, Proc. European Control Conference, July 2016, Denmark. This paper presents a lateral control strategy for a platoon of vehicles which utilises only data which can realistically be measured by each vehicle, augmented with Inter- Vehicle Communication (IVC). The control problem resembles those which exist for longitudinal control and this introduces the challenge of estimating a vehicles lateral position and velocity when direct measurement is not possible (due to lane markings being obscured by a preceding vehicle). It is shown that the associated robust controller, which we propose, exhibits string stability in the presence of sensor and actuation delays and a high fidelity simulation is conducted to verify this.
2016 Towards Formal Verification of Small and Micro UAS (2016), by S M Veres, O McAree and J M. Aitken. Proc. European Control Conference, July 2016, Denmark. The problem of verifying autonomous operations of UAS engineering systems in various types of environments such as enclosed areas, congested areas, over countryside under 400ft and in national airspace are considered. The models proposed are general enough to be applicable to most practical autonomous UAS and their subsystems. Due to the complexity of modelling and verifying a physical UAS, including experimentation in flight, the problem is split into parts of functional verification of autopilot control systems, verification of environmental perception, situational awareness, verification of goal oriented decision making of autonomous UAS and also verifying all distributed software onboard for functional correctness. The paper examines the possibilities of computer- based formal verification for each part and identifies gaps where new techniques can be useful. The paper concludes that the complete task of probabilistic verification of a UAS is well within the capabilities of today’s computing resources and could be potentially applicable in certification procedures by civil aviation authorities.
2016 Testing, Verification and Improvements of Timeliness in ROS processes (2016), by M Y Hazim, H Qu, and M. Veres, Vol 9716 of Lecture Notes in Computer Science pp 146-157
Also at TAROS 2016: Int. Conference on Towards Autonomous Robotic Systems http://link.springer.com/chapter/10.1007%2F978-3-319-40379-3_15
This paper addresses the problem of improving response times of robots implemented in the Robotic Operating System (ROS) using formal verification of computational-time feasibility. In order to verify the real time behaviour of a robot under uncertain signal processing times, methods of formal verification of timeliness properties are proposed for data flows in a ROS-based control system using Probabilistic Timed Programs (PTPs). To calculate the probability of success under certain time limits, and to demonstrate the strength of our approach, a case study is implemented for a robotic agent in terms of operational times verification using the PRISM model checker, which points to possible enhancements to the operation of the robotic agent.
2016 Autonomous Agent Behaviour Modelled in PRISM – A Case Study (2016), By R Hoffmann1 , M Ireland1, A Miller , G Norman1 ,and S M Veres, Int. Symposium on Model Checking Software SPIN 2016: Model Checking Software pp 104-110, Springer LNCS Vo. 9641. Formal verification of autonomous agents is a growing area due to the demand that autonomous systems have to be proven safe. In this paper we present an abstract definition of autonomy which can be used to model autonomous scenarios using Markov decision processes. To demonstrate the applicability of the approach we build and verify a model of an unmanned air vehicle in an exemplary autonomous scenario, utilising this approach.
2016 Stabilizability of Dynamic Coalitional Games with Transferable Utility. (2016), By M SMyrnaki, D Bauso, S M Veres. NECSYS 2016, September 8-9, 2016, Tokyo, Japan Stabilizing allocations in coalitional games with transferable utility are studied within the framework of uncertain and dynamic stochastic systems. We assume that the excess of allocations relative to the evolution of coalition values can be described by a stochastic differential equation. As main contribution we provide a feedback control law of allocations to maintain the coalition within a target set. We show that the resulting dynamics is second moment stable under our feedback linear-saturated control strategy.
2016 Virtual Spring-Damper Mesh-Based Formation Control for Spacecraft Swarms in Potential Fields (2015) By Q. Chen, SM Veres, Y Wang and Y Meng J. Guidance Control and Dynamics, Vol. 38, 3, pp 535-546. This paper derives a distributed control method based on a virtual spring-damper mesh (VSDM) for the formation control of spacecraft swarms in a gravitational potential field. This method requires no information about the absolute positions or velocities of spacecraft. Bidirectional local interaction, which provides intrinsic feedback, is used, and general connected topologies are assumed. Collision avoidance and topology switching are integrated, and the convergence of the closed-loop system is assured. Approximate expressions that predict the steady-state performance of relative position errors and control accelerations are also derived by linearization and using algebraic graph representations. The VSDM method proposed in this Note is different from the virtual spring mesh algorithm explored for the deployment of mobile sensors in [12]. One spring and one damper are combined as a connection unit for relative motion control in this study, whereas in [12], virtual dampers are separately used to decrease the absolute velocity of each agent to a stationary state.
2015

Virtual Spring-Damper Mesh-Based Formation Control for Spacecraft Swarms in Potential Fields. by Q. Chen, SM Veres, Y Wang and Y Meng, Journal of Guidance, Control, and Dynamics, Vol. 38, No. 3, March 2015, pp.535-546
http://arc.aiaa.org/doi/pdf/10.2514/1.G000569

This paper derives a distributed control method based on a virtual spring-damper mesh (VSDM) for the formation control of spacecraft swarms in a gravitational potential field. This method requires no information about the absolute positions or velocities of spacecraft. Bidirectional local interaction, which provides intrinsic feedback, is used, and general connected topologies are assumed. Collision avoidance and topology switching are integrated, and the convergence of the closed-loop system is assured. Approximate expressions that predict the steady-state performance of relative position errors and control accelerations are also derived by linearization and using algebraic graph representations.
2014

Practical Verification of Decision-Making in Agent-Based Autonomous Systems, by Louise A. Dennis, Michael Fisher, Nicholas K. Lincoln, Alexei Lisitsa, Sandor M. Veres, Automated Software Engineering, 2014, to appear. Available upon request now and will be open access later

This paper presents a verification methodology for analysing the decision-making component in agent-based hybrid systems. Traditionally hybrid automata have been used to both implement and verify such systems, but hybrid automata based modelling, programming and verification techniques scale poorly as the complexity of discrete decision-making increases making them unattractive in situations where complex logical reasoning is required. In the programming of complex systems it has, therefore, become common to separate out logical decision-making into a separate, discrete, component. However, verification techniques have failed to keep pace with this development. We are exploring agent-based logical components and have developed a model checking technique for such components which can then be composed with a separate analysis of the continuous part of the hybrid system. Among other things this allows program model checkers to be used to verify the actual implementation of the decision-making in hybrid autonomous systems.

2014

Reconfigurable Autonomy
Künstliche Intelligenz (308)
2014, to appear, DOI: 10.1007/s13218-014-0308-1

This paper outlines our reconfigurable and self-reconfiguring robotic software architectures under development. Our rational agent is able to reason about the requirements and consequences of such changes; this provides high-level description and analysis of reconfigurability. As well as reasoning about the decisions taken behind the reconfiguration it should be able to effectively communicate these decisions back to the system operators. This closes the loop around the design process but provides an important view to remove any opacity within the reconfiguration process.

2014

On Efficient Consistency Checks by Robot, (2014), by Hongyang Qu and Sandor M Veres, The European control Conference

Most autonomous robotic agents use logic inference to keep to safe and permitted behaviour. Given a set of rules, it is important that the robot is able to establish the consistency of its rules and its current perception-based beliefs. This paper investigates how a robotic agent can use model checking to examine the consistency of its rules and beliefs. A rule set is modelled by a Boolean evolution system with synchronous semantics which can be translated into a labelled transition system (LTS). It is proven that stability and consistency can be formulated as computation tree logic (CTL) and linear temporal logic (LTL) properties. Two new algorithms are presented to perform realtime consistency and stability checks respectively, which is crucial for efficient consistency checks by robots.

Programme of ECC 2014, Strasbourg, France,

2014

Formulating Robot Pursuit-Evasion Strategies by Model Checking, (2014), by Hongyang Qu, Andreas Kolling and Sandor M Veres, IFAC World Congress, Cape Town, South Africa

This paper presents an application of a model checking framework to robotic search, particularly search problems known as pursuit-evasion that assume a smart, fast and evading target. Within the framework we can model dierent pursuit-evasion problems and thereby enable a direct and rigorous comparison between different problem formulations and their respective properties and algorithms. In addition, we enable the computation of new kinds of solutions to pursuit-evasion problems, so called strategies, that can consider multiple criteria, e.g. order of vertices or connectedness. These strategies are computed by satisfying a temporal logic formula by model checking. We present theorems that show the connection between a strategy and a temporal logic formula. We demonstrate our approach by applying it to two dierent graph-based pursuit-evasion problems and show how to enable a comparison. This work presents the first step and basis for further investigations of more rigorous and unified approaches to understand, compare and design pursuit-evasion models.

Programme of IFAC World Congress

2014

Coordination of control in robot teams using game-theoretic learning, by M Smyrnakis and Sandor M Veres, IFAC World Congress, Cape Town, South Africa 2014

This paper presents a distributed decision making approach to the problem of control effort allocation to robotic team members. The objective is for a team of autonomous robots to coordinate their actions in order to efficiently complete a task. A novel controller design methodology is proposed which allows the robot team to work together based on a game theoretic learning algorithms using fictitious play and extended Kalman fllters. In particular each robot of the team predicts the other robots' planned actions while making decision to maximise its own expected reward that is dependent on the reward for joint successful completion of the task. After theoretical analysis the performance of the proposed algorithm is tested on a scenario of collaboration between material handling and patrolling robots in a warehouse.

Programme of IFAC World Congress

2014

Adaptation of System Configuration under the Robot Operating System, by Jonathan M. Aitken, Sandor M Veres and Mark Judge, IFAC World Congress, Cape Town, South Africa 2014

This paper lays down the foundations of developing a reconfigurable control system within the Robot Operating System (ROS) for autonomous robots. The essential components of robots are programmed under a ROS system. A formal model is defined as a tripartite graph to represent the robots functional architecture. ROS systems are then generalised to component libraries for any ROS architecture and an abstract model as a system graph is introduced. Orthogonality of a library and a system graph is defined and redundancy levels of robot components are studied for maintaining full functionality of the robot by automated reconfiguration in face of hardware malfunction. This allows AI planning tools, such as Planning Domain Definition Language (PDDL), to compute permissible reconfigurations. We present an example of a pair of robotic arms which requires reconfiguration of the underlying control system in order to retain the capability to carry out a task.

Programme of IFAC World Congress

2013

Autonomous Asteroid Exploration by Rational Agents, IEEE Computational Intelligence Magazine, Vol. 8, No 4, pp 25-38, 2013
DOI: 10.1109/MCI.2013.2279559

The history of software agent architectures has been driven by the parallel requirements of real-time decision-making and the sophistication of the capabilities the agent can provide. Starting from reactive, rule based, subsumption through to layered and belief-desire-intention architectures, a compromise always has to be reached between the ability to respond to the environment in a timely manner and the provision of capabilities that cope with relatively complex environments. In the spirit of these past developments, this paper is proposing a novel "anthropomorphic" agent architecture that brings together the most desirable features: natural language definitions of agent reasoning and skill descriptions, shared knowledge with operators, the combination of fast reactive as well as long term planning, ability to explain why certain actions are taken by the autonomous agent and finally inherent formal verifiability. With these attributes, the proposed agent architecture can potentially cope with the most demanding autonomous space missions.
2012

Natural Language Programming of Complex Robotic BDI Agent, by Nick K Lincoln and Sandor M Veres, Springer Journal of Intelligent and Robotic Systems © 2012,
BibTex

This paper presents a natural language design environment that enables the programming of complex robotic agent systems, comprising of a top level BDI architecture in conjunction with a low level operational system that relates to the hardware interface and supplemental computational processes. The design environment enforces synergy between the development of these traditionally disparate aspects through sharing of ontological information and implementing a form of natural language programming called sEnglish. The resultant system provides an inherent abstraction of defined operational concepts and procedures for agent reasoning and shared meaning between man and machine. Through this shared knowledge the robot’s operational logic and skill execution details are clear to human operators and may thus facilitate the work of design teams to enable rapid prototyping of physical agent systems in simulation or hardware.

EuroNews Interview - May 2011

2011

Knowledge of machines: review and forward look, by Sandor M Veres, Journal of Systems and Control Engineering, 2011
BibTex

This paper looks at the possibility of knowledge development of machines from an engineer’s point of view. First, the increasing need for autonomous operations of vehicles, manufacturing facilities, utility networks, and robots is reviewed. From the point of view of operational design, five desirable ingredients are identified that can facilitate control autonomy. Most of these facilitators can be enabled by using the methodology of natural language programming and writing of documents that machines can read and utilize to improve their feedback control skills, their knowledge of the environment, and also their decision-making skills. A forward look outlines the benefits of ‘publishing for machines’, to manufacturing, vehicle operations, utility networks, and robots.

SAGE Press Release

Machines to compare notes online? - Science Daily

Further reports

IEEE Intelligent Systems "In The News"

2011

A natural language programming solution for executable paper, by Sandor M Veres and J. Patrik Adolfsson, Procedia Computer Science ©, 2011
BibTex

The paper describes a system for executable papers for publishers enabling them to reuse content and to generate further advances of science and engineering. The executable algorithmic descriptions within a paper are presented in natural language sentences and basic code, thereby making long term compatibility absolute. Authors are required to use publicly numerical libraries on the Internet or references to publications with executable papers. As used by authors the system automatically creates a web of algorithmic knowledge on the Internet. Novelty of new algorithms in publications can be evaluated by automated tools available to authors, reviewers and readers of scientific papers published.

2011

Decision Methods for Non-Tethered Deep Space Interferometr, by N.K.Lincoln & S.M.Veres, 18th IFAC © World Congress, Milano, Italy (2011), pp. 60-65.

This paper examines decision methods applied to a distributed system that must operate collaboratively to achieve a common objective and tolerate partial system failure, necessitating structural reorganization. The problem addressed is controlled and cooperative action of multiple physical agents in a space environment, tasked with forming an interferometer, maximizing the interferometer lifetime by staged placement of resources and maintaining optimality in the presence of individual agent failures. Whilst a seemingly niche application, the methods developed and analyzed are equally valid for any multiplephysical agent system such as UAVs or AUVs. Centralized optimization methods will be applied to form decisions in addition to a distributed contract-net like decision method, wherein an auction system is adopted to form group decisions. These methods are investigated for the practicality of applying such methods to an engineering system.

2011

Hybrid Automata Discretising Agents for Formal Modelling of Robots, by L Molnar and S M Veres, 18th IFAC © World Congress, Milano, Italy, (2011), pp. 49-54.

Some of the fundamental capabilities required by autonomous vehicles and systems for their intelligent decision making are: modelling of the environment and forming data abstractions for symbolic, logic based reasoning. The paper formulates a discrete agent framework that abstracts and controls a hybrid system that is a composition of hybrid automata modelled continuous individual processes. Theoretical foundations are laid down for a class of general model composition agents (MCAs) with an advanced subclass of rational physical agents (RPAs). We define MCAs as the most basic structures for the description of complex autonomous robotic systems. The RPA’s have logic based decision making that is obtained by an extension of the hybrid systems concepts using a set of abstractions. The theory presented helps the creation of robots with reliable performance and safe operation in their environment. The paper emphasizes the abstraction aspects of the overall hybrid system that emerges from parallel composition of sets of RPAs and MCAs.

2011

Verifying fault tolerance and self-diagnosability of an autonomous underwater vehicle, by Jonathan Ezekiel, Alessio Lomuscio., Levente Molnar, Sandor Veres, Miles Pebody, 22nd International Joint Conference on Artificial Intelligence, 2011, BibTex

The paper reports on results obtained during the verification of Autosub6000, an autonomous underwater vehicle used for deep oceanic exploration. Our starting point is the Simulink/Matlab engineering model of the submarine, which is discretised by a compiler into a representation suitable for model checking. We assess the ability of the vehicle to function under degraded conditions by injecting faults automatically into the discretised model. The resulting system is analysed by means of the model checker MCMAS, and conclusions are drawn on the system’s ability to withstand faults and to perform self-diagnosis and recovery. We present lessons learnt from this and suggest a general method for verifying autonomous vehicles.

2010

Autonomous Vehicle Control Systems – a Review of Decision Makin, by S M Veres, N K Lincoln, L Molnar, C Morice, Journal of Systems and Control ©, 2010

A systematic review is provided on artificial agent methodologies applicable to control engineering of autonomous vehicles and robots. The paper focuses on some fundamentals that make a machine autonomous: decision making that involves modelling the environment and forming data abstractions for symbolic processing and logic-based reasoning. Most relevant capabilities such as navigation, autonomous path planning, path following control, and communications, that directly affect decision making, are treated as basic skills of agents. Although many autonomous vehicles have been engineered in the past without using the agent-oriented approach, most decision making onboard of vehicles is similar to or can be classified as some kind of agent architecture, even if in a naïve form. First the ANSI standard of intelligent systems is recalled then a summary of the fundamental types of possible agent architectures for autonomous vehicles are presented, starting from reactive, through layered, to advanced architectures in terms of beliefs, goals, and intentions. The review identifies some missing links between computer science results on discrete agents and engineering results of continuous world sensing, actuation, and path planning. In this context design tools for ‘abstractions programming’ are identified as needed to fill in the gap between logic-based reasoning and sensing. Finally, research is reviewed on autonomous vehicles in water, on the ground, in the air, and in space with comments on their methods of decision making. One of the main conclusions of this review is that standardization of decision making through agent architectures is desirable for the future of intelligent vehicle developments and their legal certification.
2010

Reducing Code Complexity in Hybrid Control System, by Louise. A. Dennis, Michael Fisher, Nicholas K. Lincoln, Alexei Lisitsa, Sandor M. Veres, i-SAIRAS 2010, The 10th International Symposium on Artificial Intelligence, Robotics and Automation in Space Aug. 29 – Sept. 1, (2010), Sapporo, Japan

Modern control systems are limited in their ability to react flexibly and autonomously to changing situations

by the complexity inherent in handling situations wheremany variables are present. We present an architecture based on a combination of agent programming and hybrid systems for managing high level decisions in such systems. Our preliminary case study concerns satellites maintaining geo-stationary orbits. This case study suggests that the complexity of the code of such a system increases much more slowly in the face of increasing complexity of the scenario, than in a more traditional approach based on finite state machines over controller options.

i-SAIRA 2010

2010

Mission Capable Autonomous Control Systems in the Oceans, in the Air and in Space, by Sandor M. Veres, A.Hanazawa et al. (Eds.): Brain-Inspired Info. Technology, SCI 266, pp. 1–10.

The problem of optimal degree of autonomy and minimal complexity is considered for mission capable systems in spacecrafts and unmanned aeronautical systems. Levels of autonomy and system requirements are discussed. Simple and goal oriented system architectures are summarized and the prospect of plug-and-play systems is outlined. The engineering compromise to be made between degree of autonomy and verifiability of multi-agent systems is pointed out. Three main challenges are highlighted to aid progress of future development of verifiable autonomous vehicles.

Springerlink.com © Springer-Verlag Berlin Heidelberg January 2010.

2010

Frequency domain iterative feedforward/feedback tuning for MIMO ANV, by Jian Luo, Sandor M. Veres, IFAC Automatica, Vol. 46, Issue 4, April 2010.

A new gradient estimation method is proposed that relies on efficient computation of the negative gradient of the average linear quadratic cost function completely in the frequency domain. Based on the proposed theory, a new iterative tuning method is developed to solve linear multi-input multi-output Active Noise/Vibration Control problems. Compared with published iterative tuning methods, the new method has the added advantage that the number of experiments per iteration is reduced to one. Combined with the other advantage of relatively simple controller structures, the method is suitable for realtime implementation as an adaptive controller.

2010

Satellite Control Using Rational Agent Programming, by Nicholas K Lincoln, S M Veres, Louise Dennis, Michael Fisher, Alexei Lisitsa,
IEEE AI Magazine(pdf), June 2010 (vol. 25 no. 3)

Recently, the space industry has sought to abandon large monolithic platforms in favor of multiple, smaller satellites working in teams to accomplish the task of a larger vehicle through distributed methods. This article describes ongoing work at the Universities of Liverpool and Southampton on using agent programming technology to control complex autonomous satellites.

2010

Geometric Bounding Techniques For Underwater Localisation Using Range-Only Sensor, by C Morice and S M Veres, MechE Journal of Systems and Control, 2010

This paper describes the application of geometric bounding techniques to range-only navigation of an underwater vehicle. A geometric technique is defined to obtain a position fix of an underwater vehicle using a combination of dead-reckoning navigation and acoustic measurements of range between the underwater vehicle and a global positioning system (GPS)-equipped ship. An assessment is made of the accuracy to which navigational parameters can be estimated using these methods.

2010

Application of Discrete Time Sliding Mode Control to a Spacecraft in 6Do, by N K Lincoln and S M Veres, International Journal of Control , 2010

This article presents the application of two discrete-time sliding mode controllers, developed in conjunction with a potential function guidance method, to provide control in both position and attitude for a rigid, holonomic spacecraft body using thrusters only. Identification of mass and inertial matrix parameters is also included. Both controllers are demonstrated for their effectiveness under realistic actuator constraints. One of the controllers is also implemented on hardware in a representative 5 degrees of freedom (5DoF) testbed environment to show the practical performance of the methods.

2010

Aerodynamic Parameter Estimation of an UAV Based on Extended Kalman Filter and Its Higher Order Approach, by Li Meng, Liu Li and S M Veres, 2nd IEEE International Conference on Advanced Computer Control(pdf), 2010

Aerodynamic parameter estimation provides an effective way for aerospace system modeling using measured data from flight test, especially for the purpose of developing elaborate simulation environments and control systems design of Unmanned Aerial Vehicle (UAV) with short design cycles and reduced cost. However, parameter identification of airplane dynamics is complicated because of its nonlinear identification models and the combination of noisy and biased sensor measurements. The combined difficulties mentioned above make the problem of state and parameter estimation a nonlinear filtering problem. Extended Kalman Filter (EKF) is an excellent tool for this matter with the property of recursive parameter identification and excellent filtering. The standard EKF algorithm is based on a first order approximation of system dynamics. More refined linearization techniques such as iterated EKF can be used to reduce the linearization error in the EKF for highly nonlinear systems, which leads to a theoretically better result. In this paper we concentrate on the application and comparison of EKF and iterated EKF for aerodynamic parameter estimation of a fixed wing UAV. The result shows that the two methods have been able to provide accurate estimations.

2010

A Multi-agent approach to Integrated FDI & Reconfiguration of Autonomous AGV, by Badril Abu Bakar and S.M. Veres, AIA 2010, IASTED Conf on Artificial Intelligence and Applications, Feb 15-17, Innsbruck, Austria, 2010

This work is concerned with a new type of realtime reconfigurable control systems that is based on the use of autonomous agents. To this end, two stages have been examined in the context of intelligent agent decision making; the fault detection and identification (FDI) stage and the reconfiguration stage (RC). The FDI stage detects that a fault has occurred. It then further diagnoses the situation. The RC stage follows this by adapting or changing the control architecture to accommodate the fault. The problem is to synchronize or integrate these two stages in the overall structure of a control system on real world applications. The multi-agent architecture proposed in this paper has several advantages in terms of modularity, reliability, ability to learn and achieve overall higher robustness over past ”single software” methods. Initial tests on an example system have been carried out to demonstrate this new organisation of reconfigurable control systems.

2010

Documents for Intelligent Agents in English, by SM Veres and L Molnar, AIA 2010, IASTED Conf on Artificial Intelligence and Applications, Feb 15-17, Innsbruck, Austria, 2010

The paper presents an information processing system for autonomously operating vehicles where engineers can write ”publications” written in English that the autonomoussystems can ”read” to acquire knowledge such as various skills and behaviour policies. Knowledge about how to perform feedback control based operations, how to do dynamical modelling, path planning, servo and tracking control skills, vision based feedback control, etc. can also be transferred. The ”publications” are similar to engineering booklets with contents, sections, subsections in English, that can appear in HTML, LaTex(pdf) formats. The same paper’s HTML file can be read by an agent on board the autonomous vehicle and after reading the paper the agent knows how to alter its control of the vehicle. There is no need for an application engineer to read a journal publication and implement it by programming, the research engineer’s methodological work is directly utilized by the autonomous vehicle or robot. Engineers different from the author of the ”publication” can also read the papers and learn the details of how the vehicle operates, how decisions are reached and how skills are performed. Not only will users of the autonomous vehicle clearly understand how it operates but will also know its limitations to avoid misuse or misunderstanding. Users can modify English sentences in the ”publication” to influence the vehicle’s behaviour or how its skills are performed.

2010

An Agent Based Framework for Adaptive Control and Decision Making of Autonomous Vehicles, by Nicholas K. Lincoln, Sandor M. Veres, Louise Dennis, Michael Fisher, Alexei Lisitsa, IFAC Workshop ALCOSP’2010, August 2010.

The paper addresses the problem of defining a theoretical physical agent framework that combines rational agent decision making with abstractions from predictions and planning of the future of the physical environment. The objective of the new framework is to reduce complexity of logical inference of agents controlling autonomous vehicles and robots in space exploration, deep underwater exploration, defense reconnaissance, automated manufacturing and household automation. An essential feature of the framework is automated realtime evaluations of abstractions on the effects of future actions. Comparison is made with hybrid automaton based solutions in terms of computational complexity

2010

Direct Policy Search Method in Fault Tolerant Autonomous Systems, by Abu Bakar and S.M. Veres, IFAC Workshop ALCOSP’2010, August 2010.

This work is concerned with a new type of realtime reconfigurable control systems that is based on the use of a multi-agent system. To this end, two stages have been examined in the context of decision making; the fault detection and identification (FDI) stage and the reconfiguration stage (RC). The agent based FDI detects that a fault has occurred. It then further diagnoses the situation. The RC stage follows this by adapting or changing the control architecture to accommodate the fault. The agent based problem is to synchronize or integrate these two stages in the overall structure of a control system on real world applications. The multi-agent architecture proposed in this paper has several advantages in terms of modularity, reliability, ability to learn and achieve overall higher robustness over past "single software" methods. Specifically, this paper concentrates on one of the agents introduced, namely the Reconfiguration agent. Tests on a simulation of an example system have been carried out to demonstrate this new organisation of reconfigurable control systems.

Back to top