Discrete Event Systems Based Robotic Task Modeling - Revision history

Pal: /* Institutional Controllers for Distributed Robotic Systems */

2012-05-24T13:49:11Z

Institutional Controllers for Distributed Robotic Systems

← Older revision		Revision as of 13:49, 24 May 2012
Line 20:		Line 20:

	== Institutional Controllers for Distributed Robotic Systems ==		== Institutional Controllers for Distributed Robotic Systems ==
	A formalism to represent institutions by Petri nets, within the Institutional Robotics framework, has been developed. We have assessed the ability of our formalism to replicate results obtained with other approaches in simulation and with swarms of up to 40 real e-puck robots. In particular, we have worked on a case study concerned with a swarm of simple robots which has to maintain wireless connectivity and compare results from realistic simulations performed with the institutional controller to those performed with a Finite State Automata controller. Petri net based institutional controllers can be used in modeling and analyzing the distributed robotic system they control by providing the necessary structure to build macroscopic models of that system, and studying their stochastic performance properties.		A formalism to represent institutions by Petri nets, within the [http://mediawiki.isr.ist.utl.pt/wiki/Institutional_Robotics Institutional Robotics] framework, has been developed. We have assessed the ability of our formalism to replicate results obtained with other approaches in simulation and with swarms of up to 40 real e-puck robots. In particular, we have worked on a case study concerned with a swarm of simple robots which has to maintain wireless connectivity and compare results from realistic simulations performed with the institutional controller to those performed with a Finite State Automata controller. Petri net based institutional controllers can be used in modeling and analyzing the distributed robotic system they control by providing the necessary structure to build macroscopic models of that system, and studying their stochastic performance properties.

	== Robotic task supervision using LTL ==		== Robotic task supervision using LTL ==

Pal at 13:48, 24 May 2012

2012-05-24T13:48:01Z

← Older revision		Revision as of 13:48, 24 May 2012
Line 18:		Line 18:
	<BR>		<BR>
	One may have different FSA or PN views, namely a logical untimed view, where the model can be studied based on the language (event strings) it generates, or a stochatic timed view, where time is of concern, and considered stochastically distributed. If stochastic interevent time is associated to transitions, one gets stochatic timed FSA or PN. If the event time pdf is exponential, both the FSA and the PN reachability graph can be made equivalent to Markov Chanis. If we consider the controllable event subset as subject to decisions by a higher command level, the model ends up being a Markov Decision Process (MDP), and dynamic programming or reinforcement learning solutions can be applied to determine the optimal plan in terms of minimizing some cost function, where costs are associated to primitive actions, and may include thprimitive action reliability.		One may have different FSA or PN views, namely a logical untimed view, where the model can be studied based on the language (event strings) it generates, or a stochatic timed view, where time is of concern, and considered stochastically distributed. If stochastic interevent time is associated to transitions, one gets stochatic timed FSA or PN. If the event time pdf is exponential, both the FSA and the PN reachability graph can be made equivalent to Markov Chanis. If we consider the controllable event subset as subject to decisions by a higher command level, the model ends up being a Markov Decision Process (MDP), and dynamic programming or reinforcement learning solutions can be applied to determine the optimal plan in terms of minimizing some cost function, where costs are associated to primitive actions, and may include thprimitive action reliability.

			== Institutional Controllers for Distributed Robotic Systems ==
			A formalism to represent institutions by Petri nets, within the Institutional Robotics framework, has been developed. We have assessed the ability of our formalism to replicate results obtained with other approaches in simulation and with swarms of up to 40 real e-puck robots. In particular, we have worked on a case study concerned with a swarm of simple robots which has to maintain wireless connectivity and compare results from realistic simulations performed with the institutional controller to those performed with a Finite State Automata controller. Petri net based institutional controllers can be used in modeling and analyzing the distributed robotic system they control by providing the necessary structure to build macroscopic models of that system, and studying their stochastic performance properties.

	== Robotic task supervision using LTL ==		== Robotic task supervision using LTL ==
Line 144:		Line 147:
	<BR>		<BR>
	[13] Bruno Lacerda, Pedro Lima, "Designing Petri Net Supervisors from LTL Specifications", Proc. of the 2011 Robotics: Science and Systems Conference, Los Angeles, CA, USA, 2011		[13] Bruno Lacerda, Pedro Lima, "Designing Petri Net Supervisors from LTL Specifications", Proc. of the 2011 Robotics: Science and Systems Conference, Los Angeles, CA, USA, 2011
			<BR>
			[14] José N. Pereira, Porfírio Silva, Pedro Lima, A. Martinoli, "Formalizing Institutions as Executable Petri Nets for Distributed Robotic Systems", Proc. of ECAL 2011 - European Conference on Artificial Life, Paris, France, 2011

Pal: /* References */

2012-05-24T13:41:16Z

References

← Older revision		Revision as of 13:41, 24 May 2012
Line 140:		Line 140:
	<BR>		<BR>
	[11] [http://www.springerlink.com/content/g4q46786645730n4/ Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous Robots, 2012, DOI 10.1007/s10514-012-9288-x, 2012]		[11] [http://www.springerlink.com/content/g4q46786645730n4/ Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous Robots, 2012, DOI 10.1007/s10514-012-9288-x, 2012]
			<BR>
	[12] Bruno Lacerda, Pedro Lima, "LTL-Based Decentralized Supervisory Control of Multi-Robot Tasks Modelled as Petri Nets", Proc. of IROS 2011 - IEEE/RSJ International Conference on Intelligent Robots and Systems, San Francisco, CA, USA, 2011		[12] Bruno Lacerda, Pedro Lima, "LTL-Based Decentralized Supervisory Control of Multi-Robot Tasks Modelled as Petri Nets", Proc. of IROS 2011 - IEEE/RSJ International Conference on Intelligent Robots and Systems, San Francisco, CA, USA, 2011
			<BR>
	[13] Bruno Lacerda, Pedro Lima, "Designing Petri Net Supervisors from LTL Specifications", Proc. of the 2011 Robotics: Science and Systems Conference, Los Angeles, CA, USA, 2011		[13] Bruno Lacerda, Pedro Lima, "Designing Petri Net Supervisors from LTL Specifications", Proc. of the 2011 Robotics: Science and Systems Conference, Los Angeles, CA, USA, 2011

Pal: /* References */

2012-05-24T13:40:46Z

References

← Older revision		Revision as of 13:40, 24 May 2012
Line 137:		Line 137:
	[9] D. Milutinovic, J. Carneiro, M. Athans, P. Lima, "Modeling Dynamics of Cell Population Molecule Expression Distribution", Journal of Non-Linear Analysis: Hybrid Systems and Applications, Elsevier, Vol. 1, Issue 1, pp. 81-94, 2007		[9] D. Milutinovic, J. Carneiro, M. Athans, P. Lima, "Modeling Dynamics of Cell Population Molecule Expression Distribution", Journal of Non-Linear Analysis: Hybrid Systems and Applications, Elsevier, Vol. 1, Issue 1, pp. 81-94, 2007
	<BR>		<BR>
	[10] D. Milutinovic, P. Lima, "Cells and Robots - Modeling and Control of Large-Size Agent Populations", Springer Tracts in Advanced Robotics (STAR) Series , Vol. 32, 2007		[10] [http://www.springer.com/engineering/robotics/book/978-3-540-71981-6 D. Milutinovic, P. Lima, "Cells and Robots - Modeling and Control of Large-Size Agent Populations", Springer Tracts in Advanced Robotics (STAR) Series , Vol. 32, 2007]
	<BR>		<BR>
	[11] Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous Robots, 2012, DOI 10.1007/s10514-012-9288-x, 2012		[11] [http://www.springerlink.com/content/g4q46786645730n4/ Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous Robots, 2012, DOI 10.1007/s10514-012-9288-x, 2012]
			[12] Bruno Lacerda, Pedro Lima, "LTL-Based Decentralized Supervisory Control of Multi-Robot Tasks Modelled as Petri Nets", Proc. of IROS 2011 - IEEE/RSJ International Conference on Intelligent Robots and Systems, San Francisco, CA, USA, 2011
			[13] Bruno Lacerda, Pedro Lima, "Designing Petri Net Supervisors from LTL Specifications", Proc. of the 2011 Robotics: Science and Systems Conference, Los Angeles, CA, USA, 2011

Pal: /* References */

2012-05-23T18:38:43Z

References

← Older revision		Revision as of 18:38, 23 May 2012
Line 139:		Line 139:
	[10] D. Milutinovic, P. Lima, "Cells and Robots - Modeling and Control of Large-Size Agent Populations", Springer Tracts in Advanced Robotics (STAR) Series , Vol. 32, 2007		[10] D. Milutinovic, P. Lima, "Cells and Robots - Modeling and Control of Large-Size Agent Populations", Springer Tracts in Advanced Robotics (STAR) Series , Vol. 32, 2007
	<BR>		<BR>
	[11] Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous ~~rOBOTS~~, 2012, DOI 10.1007/s10514-012-9288-x, 2012		[11] Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous Robots, 2012, DOI 10.1007/s10514-012-9288-x, 2012

Pal: /* References */

2012-05-23T18:38:19Z

References

← Older revision		Revision as of 18:38, 23 May 2012
Line 138:		Line 138:
	<BR>		<BR>
	[10] D. Milutinovic, P. Lima, "Cells and Robots - Modeling and Control of Large-Size Agent Populations", Springer Tracts in Advanced Robotics (STAR) Series , Vol. 32, 2007		[10] D. Milutinovic, P. Lima, "Cells and Robots - Modeling and Control of Large-Size Agent Populations", Springer Tracts in Advanced Robotics (STAR) Series , Vol. 32, 2007
			<BR>
			[11] Hugo Costelha, Pedro Lima, "Robot task plan representation by Petri nets: modelling, identification, analysis and execution", Autonomous rOBOTS, 2012, DOI 10.1007/s10514-012-9288-x, 2012

Pal: /* Stochastic Hybrid Automaton Model of Large Robot Populations */

2008-11-24T16:37:40Z

Stochastic Hybrid Automaton Model of Large Robot Populations

← Older revision		Revision as of 16:37, 24 November 2008
Line 80:		Line 80:
	A similar concept was applied to the ISocRob 4LL team, with 2 AIBOs exchanging ball passes - see [http://islab.isr.ist.utl.pt/wikiPicsVideos/firstPass_withPN.avi video, including simultaneous PN execution]. This was joint work with Daniele Nardi's group, at U. Rome "La Sapienza" [6][7].		A similar concept was applied to the ISocRob 4LL team, with 2 AIBOs exchanging ball passes - see [http://islab.isr.ist.utl.pt/wikiPicsVideos/firstPass_withPN.avi video, including simultaneous PN execution]. This was joint work with Daniele Nardi's group, at U. Rome "La Sapienza" [6][7].

	== Stochastic ~~Hybrid Automaton Model~~ of ~~Large Robot Populations~~ ==		== Stochastic hybrid automata model of large robot populations ==
	We have introduced a novel approach to the study of the dynamics of		We have introduced a novel approach to the study of the dynamics of
	molecule expression level in large-size cell populations, whose goal is to		molecule expression level in large-size cell populations, whose goal is to

Pal: /* Stochastic Hybrid Automaton Model of Large Robot Populations */

2008-11-23T11:15:27Z

Stochastic Hybrid Automaton Model of Large Robot Populations

← Older revision		Revision as of 11:15, 23 November 2008
Line 100:		Line 100:
	is modeled by a partial differential equation. This work was motivated by the investigation		is modeled by a partial differential equation. This work was motivated by the investigation
	of T-cell receptor expression distribution. These receptors are essential for the antigen		of T-cell receptor expression distribution. These receptors are essential for the antigen
	recognition and the regulation of the immune system.		recognition and the regulation of the immune system. [9]
	<BR>		<BR>
	For robot/agent populations, this resulted in an approach to the modeling		For robot/agent populations, this resulted in an approach to the modeling
Line 109:		Line 109:
	optimal control of partial differential equations was exploited to		optimal control of partial differential equations was exploited to
	solve this problem and it was applied to the mission control of a		solve this problem and it was applied to the mission control of a
	simulated large robotic population.		simulated large robotic population. [8]
	<BR>		<BR>
	This work was mainly the result of Dejan Milutinovic's PhD thesis, published in a Springer STAR Series book [10].		This work was mainly the result of Dejan Milutinovic's PhD thesis, published in a Springer STAR Series book [10].

Pal: /* Stochastic Hybrid Automaton Model of Large Robot Populations */

2008-11-23T11:14:43Z

Stochastic Hybrid Automaton Model of Large Robot Populations

← Older revision		Revision as of 11:14, 23 November 2008
Line 113:		Line 113:
	This work was mainly the result of Dejan Milutinovic's PhD thesis, published in a Springer STAR Series book [10].		This work was mainly the result of Dejan Milutinovic's PhD thesis, published in a Springer STAR Series book [10].
	[[Image:DejanBookCover.jpg\|thumb‎\|50px]]		[[Image:DejanBookCover.jpg\|thumb‎\|50px]]
			<BR>
			Prof. Michael Athans (Emeritus Professor at MIT) was also involved in the co-supervision of Dejan, and this work
			started up a collaboration with [http://www.igc.gulbenkian.pt/ Instituto Gulbenkian de Ciência], namely with [http://eao.igc.gulbenkian.pt/ti/Gente/Jorge/index.pt.html Prof. Jorge Carneiro], coordinator of the Theoretical Immunology group.

	== References ==		== References ==

Pal: /* Stochastic Hybrid Automaton Model of Large Robot Populations */

2008-11-23T11:11:59Z

Stochastic Hybrid Automaton Model of Large Robot Populations

← Older revision		Revision as of 11:11, 23 November 2008
Line 112:		Line 112:
	<BR>		<BR>
	This work was mainly the result of Dejan Milutinovic's PhD thesis, published in a Springer STAR Series book [10].		This work was mainly the result of Dejan Milutinovic's PhD thesis, published in a Springer STAR Series book [10].
	[[Image:DejanBookCover.jpg\|thumb‎\|~~10pk~~]]		[[Image:DejanBookCover.jpg\|thumb‎\|50px]]

	== References ==		== References ==