Roman Bartak - Academia.edu (original) (raw)

Papers by Roman Bartak

Synthesis Lectures on Artificial Intelligence and Machine Learning, 2014

Solving challenging computational problems involving time has been a critical component in the de... more Solving challenging computational problems involving time has been a critical component in the development of artificial intelligence systems almost since the inception of the field. is book provides a concise introduction to the core computational elements of temporal reasoning for use in AI systems for planning and scheduling, as well as systems that extract temporal information from data. It presents a survey of temporal frameworks based on constraints, both qualitative and quantitative, as well as of major temporal consistency techniques. e book also introduces the reader to more recent extensions to the core model that allow AI systems to explicitly represent temporal preferences and temporal uncertainty. is book is intended for students and researchers interested in constraint-based temporal reasoning. It provides a self-contained guide to the different representations of time, as well as examples of recent applications of time in AI systems.

Proceedings of the International Symposium on Combinatorial Search

Solving planning problems via translation to propositional satisfiability (SAT) is one of the mos... more Solving planning problems via translation to propositional satisfiability (SAT) is one of the most successful approaches to automated planning. An important aspect of this approach is the encoding, i.e., the construction of a propositional formula from a given planning problem instance. Numerous encoding schemes have been proposed in the recent years each aiming to outperform the previous encodings on the majority of the benchmark problems. In this paper we take a different approach. Instead of trying to develop a new encoding that is better for all kinds of benchmarks we take recently developed specialized encoding schemes and design a method to automatically select the proper encoding for a given planning problem instance. In the paper we also examine ranking heuristics for the Relaxed Relaxed Exists-Step encoding, which plays an important role in our algorithm. Experiments show that our new approach significantly outperforms the state-of-the-art encoding schemes when compared on ...

Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment

With more computing power available, video games may spare increasing amounts of processing time ... more With more computing power available, video games may spare increasing amounts of processing time for AI. One prospective application of the newly available resources is the simulation of large amounts of non-player characters (NPCs) in open world games. While it is relatively easy to simulate simple behaviours of individual NPCs it is much more difficult to create meaningful interactions between the NPCs. However, without interaction, the world cannot look very alive. In this paper we present a technique that enriches the NPC simulation with pre-scripted situations - short sketches involving coordinated interaction between several NPCs that do not substantially alter the state of the game world but increase the appeal of the world to the player. We use constraint satisfaction techniques to find NPCs suitable to enact the situations at runtime. We have implemented situations on top of the AI system for an upcoming AAA open-world game and show that this approach satisfies functional a...

Abstract. 1 The conventional wisdom and practice is that planning and scheduling tasks are solved... more Abstract. 1 The conventional wisdom and practice is that planning and scheduling tasks are solved separately using different methods and approaches. However, recent development in industrial planning and scheduling demands for mixing both tasks to allow modelling of wider class of problems. The purpose of this paper is to present a framework for mixing planning and scheduling tasks within single system. We analyse traditional views of planning and scheduling and we highlight the drawbacks of separating both tasks when applied to modelling complex process environments. We give some real-life examples where the mixed approach helps to model the problems and we propose a generic framework for such mixture. We also argue for using constraint programming as underlying solving technology and, finally, we describe some constraint models based on the proposed framework. Although, we concentrate on planning and scheduling in complex process environments we believe that the results contribute...

Problem Areas In general, the VisOpt scheduler is appropriate for almost all production schedulin... more Problem Areas In general, the VisOpt scheduler is appropriate for almost all production scheduling problems. In particular, we concentrate on problems where many different types of resources appear. We call such problems heterogeneous because the resources involved in the scheduling task can be of many different types, like producer, mover, store, tool, worker, etc. Actually, we include custom and purchase orders among resources because, in our terminology, resource is everything that is being scheduled. Note that the scheduling engine is (almost) independent of particular structure of resources so it is possible to extend the scheduler with other resource types.

Proceedings of the 12th International Conference on Agents and Artificial Intelligence, 2020

In this paper, we study the problem of an intelligent intersection. There are many studies that p... more In this paper, we study the problem of an intelligent intersection. There are many studies that present an algorithm that tries to efficiently coordinate many agents in a given intersection, however, in this paper, we study the layout of the intersection and its implications to the quality of the plan. We start with two of today commonly used road intersections (4-way intersection and roundabout) and compare them with an intersection that is less restrictive on the movements of the agents. This means that the agents do not have to use predefined lanes or follow a prescribed driving direction. We also study the effect of granularity of the intersection. The navigation of the agents in a given intersection is solved as an instance of (online) multi-agent path-finding problem.

Proceedings of 13th International Workshop on Advanced Computing and Analysis Techniques in Physics Research — PoS(ACAT2010), 2011

Building Efficient Data Planner for Peta-scale Science Michal ZEROLA Figure 1: General view of th... more Building Efficient Data Planner for Peta-scale Science Michal ZEROLA Figure 1: General view of the automated planning system. The goal is to achieve controlled and efficient utilization of the network and data services with a proper use of existing point-to-point transfer tools. At the highest level of abstraction, the planner should appear as a "box" between the user's requests and the resources.

Acta Polytechnica CTU Proceedings, 2015

Solving planning problems via translation to satisfiability (SAT) is one of the most successful a... more Solving planning problems via translation to satisfiability (SAT) is one of the most successful approaches to automated planning. We propose a new encoding scheme, called Reinforced Encoding, which encodes a planning problem represented in the SAS+ formalism into SAT. The Reinforced Encoding is a combination of the transition-based SASE encoding with the classical propositional encoding. In our experiments we compare our new encoding to other known SAS+ based encodings. The results indicate, that he Reinforced encoding performs well on the benchmark problems of the 2011 International Planning Competition and can outperform all the other known encodings for several domains.

Service level agreements (SLAs) are powerful instruments for describing all obligations and expec... more Service level agreements (SLAs) are powerful instruments for describing all obligations and expectations in a business relationship. It is of focal importance for deploying Grid technology to commercial applications. The EC-funded project HPC4U (Highly Predictable Clusters for Internet Grids) aimed at introducing SLA-awareness in local resource management systems, while the EC-funded project AssessGrid introduced the notion of risk, which is associated with every business contract. This paper highlights the concept of planning based resource management and describes the SLA-aware scheduler developed and used in these projects.

Traditional planning deals with the problem of finding activities to satisfy a given goal while t... more Traditional planning deals with the problem of finding activities to satisfy a given goal while traditional scheduling solves the problem of allocating known activities to limited resources and to limited time. In many real-life problems both tasks should be accomplished together. We give an example of such integrated planning and scheduling problem and in the software demonstration we show how the problem is solved by the Visopt ShopFloor system Keywords: applications of planning and scheduling, planning and scheduling with complex domain models, constraint reasoning for planning and scheduling

Real-life problems typically include both planning and scheduling components. In some areas, like... more Real-life problems typically include both planning and scheduling components. In some areas, like manufacturing, the planning task is not very complicated but what makes the problem hard is integration with scheduling. Traditional schedulers cannot handle such problems because of the planning component and traditional planners are not interested because planning is easy there. The paper describes some sub-problems in manufacturing scheduling that can be seen as a challenge for integrated planning and scheduling.

In the paper we propose a framework describing constraint hierarchy solvers and thus providing a ... more In the paper we propose a framework describing constraint hierarchy solvers and thus providing a theoretical foundation for construction of such solvers. The framework is based on the decomposition of the constraint hierarchy into constraint cells, i.e., chunks of constraints that are solved together. Using various decompositions we can describe a scale of constraint hierarchy solvers from the refining method to the local propagation methods. We highlight important features of the decompositions that form sufficient conditions in soundness and completeness theorems.

Planning and scheduling are close areas but on the other hand, they use very different solving te... more Planning and scheduling are close areas but on the other hand, they use very different solving technologies. Recently, there is a strong demand from industry to integrate both approaches into a single system. In the paper we propose a basic framework for modelling planning and scheduling problems that involve reasoning about time and resources. In this framework we go beyond the traditional definitions of planning and scheduling and, from the beginning, we expect integration of both these areas.

1 The conventional wisdom and practice is that planning and scheduling tasks are solved separatel... more 1 The conventional wisdom and practice is that planning and scheduling tasks are solved separately using different methods and approaches. However, recent development in industrial planning and scheduling demands for mixing both tasks to allow modelling of wider class of problems. The purpose of this paper is to present a framework for mixing planning and scheduling tasks within single system. We analyse traditional views of planning and scheduling and we highlight the drawbacks of separating both tasks when applied to modelling complex process environments. We give some real-life examples where the mixed approach helps to model the problems and we propose a generic framework for such mixture. We also argue for using constraint programming as underlying solving technology and, finally, we describe some constraint models based on the proposed framework. Although, we concentrate on planning and scheduling in complex process environments we believe that the results contribute to both planning and scheduling communities in general.

Lecture Notes in Computer Science, 2005

Many real-life problems are dynamic, with changes in the problem definition occurring after a sol... more Many real-life problems are dynamic, with changes in the problem definition occurring after a solution to the initial formulation has been reached. A minimal perturbation problem incorporates these changes, along with the initial solution, as a new problem whose solution must be as close as possible to the initial solution. A new iterative forward search algorithm is proposed to solve minimal perturbation problems. Significant improvements to the solution quality are achieved by including new conflict-based statistics in this algorithm. The proposed methods were applied to find a new solution to an existing large scale class timetabling problem at Purdue University, incorporating the initial solution and additional input changes.

The Florida AI Research Society Conference, 2007

Temporal Networks play an important role in solving planning problems and they are also used, tho... more Temporal Networks play an important role in solving planning problems and they are also used, though not as frequently, when solving scheduling problems. In this paper we propose an extension of temporal networks by parallel and alternative branching. This extension supports modelling of alternative paths in the network; in particular, it is motivated by modelling alternative process routes in manufacturing

Computer Science and Software Engineering, 2014

1 The conventional wisdom and practice is that planning and scheduling tasks are solved separatel... more 1 The conventional wisdom and practice is that planning and scheduling tasks are solved separately using different methods and approaches. However, recent development in industrial planning and scheduling demands for mixing both tasks to allow modelling of wider class of problems. The purpose of this paper is to present a framework for mixing planning and scheduling tasks within single system. We analyse traditional views of planning and scheduling and we highlight the drawbacks of separating both tasks when applied to modelling complex process environments. We give some real-life examples where the mixed approach helps to model the problems and we propose a generic framework for such mixture. We also argue for using constraint programming as the underlying solving technology and, finally, we describe some constraint models based on the proposed framework. Although, we concentrate on planning and scheduling in complex process environments we believe that the results contribute to both planning and scheduling communities in general.

Journal of Intelligent Manufacturing, 2008

Synthesis Lectures on Artificial Intelligence and Machine Learning, 2014

Solving challenging computational problems involving time has been a critical component in the de... more Solving challenging computational problems involving time has been a critical component in the development of artificial intelligence systems almost since the inception of the field. is book provides a concise introduction to the core computational elements of temporal reasoning for use in AI systems for planning and scheduling, as well as systems that extract temporal information from data. It presents a survey of temporal frameworks based on constraints, both qualitative and quantitative, as well as of major temporal consistency techniques. e book also introduces the reader to more recent extensions to the core model that allow AI systems to explicitly represent temporal preferences and temporal uncertainty. is book is intended for students and researchers interested in constraint-based temporal reasoning. It provides a self-contained guide to the different representations of time, as well as examples of recent applications of time in AI systems.

Proceedings of the International Symposium on Combinatorial Search

Solving planning problems via translation to propositional satisfiability (SAT) is one of the mos... more Solving planning problems via translation to propositional satisfiability (SAT) is one of the most successful approaches to automated planning. An important aspect of this approach is the encoding, i.e., the construction of a propositional formula from a given planning problem instance. Numerous encoding schemes have been proposed in the recent years each aiming to outperform the previous encodings on the majority of the benchmark problems. In this paper we take a different approach. Instead of trying to develop a new encoding that is better for all kinds of benchmarks we take recently developed specialized encoding schemes and design a method to automatically select the proper encoding for a given planning problem instance. In the paper we also examine ranking heuristics for the Relaxed Relaxed Exists-Step encoding, which plays an important role in our algorithm. Experiments show that our new approach significantly outperforms the state-of-the-art encoding schemes when compared on ...

Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment

With more computing power available, video games may spare increasing amounts of processing time ... more With more computing power available, video games may spare increasing amounts of processing time for AI. One prospective application of the newly available resources is the simulation of large amounts of non-player characters (NPCs) in open world games. While it is relatively easy to simulate simple behaviours of individual NPCs it is much more difficult to create meaningful interactions between the NPCs. However, without interaction, the world cannot look very alive. In this paper we present a technique that enriches the NPC simulation with pre-scripted situations - short sketches involving coordinated interaction between several NPCs that do not substantially alter the state of the game world but increase the appeal of the world to the player. We use constraint satisfaction techniques to find NPCs suitable to enact the situations at runtime. We have implemented situations on top of the AI system for an upcoming AAA open-world game and show that this approach satisfies functional a...

Abstract. 1 The conventional wisdom and practice is that planning and scheduling tasks are solved... more Abstract. 1 The conventional wisdom and practice is that planning and scheduling tasks are solved separately using different methods and approaches. However, recent development in industrial planning and scheduling demands for mixing both tasks to allow modelling of wider class of problems. The purpose of this paper is to present a framework for mixing planning and scheduling tasks within single system. We analyse traditional views of planning and scheduling and we highlight the drawbacks of separating both tasks when applied to modelling complex process environments. We give some real-life examples where the mixed approach helps to model the problems and we propose a generic framework for such mixture. We also argue for using constraint programming as underlying solving technology and, finally, we describe some constraint models based on the proposed framework. Although, we concentrate on planning and scheduling in complex process environments we believe that the results contribute...

Problem Areas In general, the VisOpt scheduler is appropriate for almost all production schedulin... more Problem Areas In general, the VisOpt scheduler is appropriate for almost all production scheduling problems. In particular, we concentrate on problems where many different types of resources appear. We call such problems heterogeneous because the resources involved in the scheduling task can be of many different types, like producer, mover, store, tool, worker, etc. Actually, we include custom and purchase orders among resources because, in our terminology, resource is everything that is being scheduled. Note that the scheduling engine is (almost) independent of particular structure of resources so it is possible to extend the scheduler with other resource types.

Proceedings of the 12th International Conference on Agents and Artificial Intelligence, 2020

In this paper, we study the problem of an intelligent intersection. There are many studies that p... more In this paper, we study the problem of an intelligent intersection. There are many studies that present an algorithm that tries to efficiently coordinate many agents in a given intersection, however, in this paper, we study the layout of the intersection and its implications to the quality of the plan. We start with two of today commonly used road intersections (4-way intersection and roundabout) and compare them with an intersection that is less restrictive on the movements of the agents. This means that the agents do not have to use predefined lanes or follow a prescribed driving direction. We also study the effect of granularity of the intersection. The navigation of the agents in a given intersection is solved as an instance of (online) multi-agent path-finding problem.

Proceedings of 13th International Workshop on Advanced Computing and Analysis Techniques in Physics Research — PoS(ACAT2010), 2011

Building Efficient Data Planner for Peta-scale Science Michal ZEROLA Figure 1: General view of th... more Building Efficient Data Planner for Peta-scale Science Michal ZEROLA Figure 1: General view of the automated planning system. The goal is to achieve controlled and efficient utilization of the network and data services with a proper use of existing point-to-point transfer tools. At the highest level of abstraction, the planner should appear as a "box" between the user's requests and the resources.

Acta Polytechnica CTU Proceedings, 2015

Solving planning problems via translation to satisfiability (SAT) is one of the most successful a... more Solving planning problems via translation to satisfiability (SAT) is one of the most successful approaches to automated planning. We propose a new encoding scheme, called Reinforced Encoding, which encodes a planning problem represented in the SAS+ formalism into SAT. The Reinforced Encoding is a combination of the transition-based SASE encoding with the classical propositional encoding. In our experiments we compare our new encoding to other known SAS+ based encodings. The results indicate, that he Reinforced encoding performs well on the benchmark problems of the 2011 International Planning Competition and can outperform all the other known encodings for several domains.

Service level agreements (SLAs) are powerful instruments for describing all obligations and expec... more Service level agreements (SLAs) are powerful instruments for describing all obligations and expectations in a business relationship. It is of focal importance for deploying Grid technology to commercial applications. The EC-funded project HPC4U (Highly Predictable Clusters for Internet Grids) aimed at introducing SLA-awareness in local resource management systems, while the EC-funded project AssessGrid introduced the notion of risk, which is associated with every business contract. This paper highlights the concept of planning based resource management and describes the SLA-aware scheduler developed and used in these projects.

Traditional planning deals with the problem of finding activities to satisfy a given goal while t... more Traditional planning deals with the problem of finding activities to satisfy a given goal while traditional scheduling solves the problem of allocating known activities to limited resources and to limited time. In many real-life problems both tasks should be accomplished together. We give an example of such integrated planning and scheduling problem and in the software demonstration we show how the problem is solved by the Visopt ShopFloor system Keywords: applications of planning and scheduling, planning and scheduling with complex domain models, constraint reasoning for planning and scheduling

Real-life problems typically include both planning and scheduling components. In some areas, like... more Real-life problems typically include both planning and scheduling components. In some areas, like manufacturing, the planning task is not very complicated but what makes the problem hard is integration with scheduling. Traditional schedulers cannot handle such problems because of the planning component and traditional planners are not interested because planning is easy there. The paper describes some sub-problems in manufacturing scheduling that can be seen as a challenge for integrated planning and scheduling.

In the paper we propose a framework describing constraint hierarchy solvers and thus providing a ... more In the paper we propose a framework describing constraint hierarchy solvers and thus providing a theoretical foundation for construction of such solvers. The framework is based on the decomposition of the constraint hierarchy into constraint cells, i.e., chunks of constraints that are solved together. Using various decompositions we can describe a scale of constraint hierarchy solvers from the refining method to the local propagation methods. We highlight important features of the decompositions that form sufficient conditions in soundness and completeness theorems.

Planning and scheduling are close areas but on the other hand, they use very different solving te... more Planning and scheduling are close areas but on the other hand, they use very different solving technologies. Recently, there is a strong demand from industry to integrate both approaches into a single system. In the paper we propose a basic framework for modelling planning and scheduling problems that involve reasoning about time and resources. In this framework we go beyond the traditional definitions of planning and scheduling and, from the beginning, we expect integration of both these areas.

1 The conventional wisdom and practice is that planning and scheduling tasks are solved separatel... more 1 The conventional wisdom and practice is that planning and scheduling tasks are solved separately using different methods and approaches. However, recent development in industrial planning and scheduling demands for mixing both tasks to allow modelling of wider class of problems. The purpose of this paper is to present a framework for mixing planning and scheduling tasks within single system. We analyse traditional views of planning and scheduling and we highlight the drawbacks of separating both tasks when applied to modelling complex process environments. We give some real-life examples where the mixed approach helps to model the problems and we propose a generic framework for such mixture. We also argue for using constraint programming as underlying solving technology and, finally, we describe some constraint models based on the proposed framework. Although, we concentrate on planning and scheduling in complex process environments we believe that the results contribute to both planning and scheduling communities in general.

Lecture Notes in Computer Science, 2005

Many real-life problems are dynamic, with changes in the problem definition occurring after a sol... more Many real-life problems are dynamic, with changes in the problem definition occurring after a solution to the initial formulation has been reached. A minimal perturbation problem incorporates these changes, along with the initial solution, as a new problem whose solution must be as close as possible to the initial solution. A new iterative forward search algorithm is proposed to solve minimal perturbation problems. Significant improvements to the solution quality are achieved by including new conflict-based statistics in this algorithm. The proposed methods were applied to find a new solution to an existing large scale class timetabling problem at Purdue University, incorporating the initial solution and additional input changes.

The Florida AI Research Society Conference, 2007

Temporal Networks play an important role in solving planning problems and they are also used, tho... more Temporal Networks play an important role in solving planning problems and they are also used, though not as frequently, when solving scheduling problems. In this paper we propose an extension of temporal networks by parallel and alternative branching. This extension supports modelling of alternative paths in the network; in particular, it is motivated by modelling alternative process routes in manufacturing

Computer Science and Software Engineering, 2014

1 The conventional wisdom and practice is that planning and scheduling tasks are solved separatel... more 1 The conventional wisdom and practice is that planning and scheduling tasks are solved separately using different methods and approaches. However, recent development in industrial planning and scheduling demands for mixing both tasks to allow modelling of wider class of problems. The purpose of this paper is to present a framework for mixing planning and scheduling tasks within single system. We analyse traditional views of planning and scheduling and we highlight the drawbacks of separating both tasks when applied to modelling complex process environments. We give some real-life examples where the mixed approach helps to model the problems and we propose a generic framework for such mixture. We also argue for using constraint programming as the underlying solving technology and, finally, we describe some constraint models based on the proposed framework. Although, we concentrate on planning and scheduling in complex process environments we believe that the results contribute to both planning and scheduling communities in general.

Journal of Intelligent Manufacturing, 2008