Results 1  10
of
40
Exact Solutions to TimeDependent MDPs
 in Advances in Neural Information Processing Systems
, 2000
"... We describe an extension of the Markov decision process model in which a continuous time dimension is included in the state space. This allows for the representation and exact solution of a wide range of problems in which transitions or rewards vary over time. We examine problems based on route ..."
Abstract

Cited by 69 (6 self)
 Add to MetaCart
(Show Context)
We describe an extension of the Markov decision process model in which a continuous time dimension is included in the state space. This allows for the representation and exact solution of a wide range of problems in which transitions or rewards vary over time. We examine problems based on route planning with public transportation and telescope observation scheduling. 1
DecisionTheoretic Control of Planetary Rovers
 Lecture Notes in Computer Science
, 2002
"... Planetary rovers are small unmanned vehicles equipped with cameras and a variety of sensors used for scientific experiments. They must operate under tight constraints over such resources as operation time, power, storage capacity, and communication bandwidth. Moreover, the limited computational r ..."
Abstract

Cited by 17 (3 self)
 Add to MetaCart
(Show Context)
Planetary rovers are small unmanned vehicles equipped with cameras and a variety of sensors used for scientific experiments. They must operate under tight constraints over such resources as operation time, power, storage capacity, and communication bandwidth. Moreover, the limited computational resources of the rover limit the complexity of online planning and scheduling. We describe two decisiontheoretic approaches to maximize the productivity of planetary rovers: one based on adaptive planning and the other on hierarchical reinforcement learning. Both approaches map the problem into a Markov decision problem and attempt to solve a large part of the problem offline, exploiting the structure of the plan and independence between plan components. We examine the advantages and limitations of these techniques and their scalability.
MissionDirected Path Planning for Planetary Rover Exploration
, 2004
"... Robotic rovers uniquely benefit planetary exploration they enable regional exploration with the precision of insitu measurements, a combination impossible from an orbiting spacecraft or fixed lander. Current rover mission planning activities utilize sophisticated software for activity planning and ..."
Abstract

Cited by 16 (1 self)
 Add to MetaCart
(Show Context)
Robotic rovers uniquely benefit planetary exploration they enable regional exploration with the precision of insitu measurements, a combination impossible from an orbiting spacecraft or fixed lander. Current rover mission planning activities utilize sophisticated software for activity planning and scheduling, but simplified path planning and execution approaches tailored for localized operations to individual targets. Routes are coarsely handselected by human operators and executed by the rover’s local obstacle detection and avoidance software. Neither route selection nor navigation deeply considers high level mission goals, large scale terrain, time, resources or operational constraints. This strategy is insufficient for the investigation of multiple, regionally distributed targets in a single command cycle. Path planning tailored for this task must consider the impact of large scale terrain on power, speed and regional access; the effect of route timing on resource availability; the limitations of finite resource capacity and other operational constraints on vehicle range and timing; and the mutual influence between traverses and upstream and downstream stationary activities. Encapsulating this reasoning in an efficient autonomous planner would allow a rover to continue operating rationally despite significant deviations from an initial plan. This research presents missiondirected path planning that enables an autonomous, strategic reasoning capability for robotic explorers. Planning operates in a space of position, time and energy. Unlike previous hierarchical
An axiomatic approach to robustness in search problems with multiple scenarios
 In Proc. of the 19th conference on Uncertainty in Artificial Intelligence. Acapulco
, 2003
"... This paper is devoted to the the search of robust solutions in state space graphs when costs depend on scenarios. We first present axiomatic requirements for preference compatibility with the intuitive idea of robustness. This leads us to propose the Lorenz dominance rule as a basis for robustness a ..."
Abstract

Cited by 15 (7 self)
 Add to MetaCart
(Show Context)
This paper is devoted to the the search of robust solutions in state space graphs when costs depend on scenarios. We first present axiomatic requirements for preference compatibility with the intuitive idea of robustness. This leads us to propose the Lorenz dominance rule as a basis for robustness analysis. Then, after presenting complexity results about the determination of robust solutions, we propose a new sophistication of A ∗ specially designed to determine the set of robust paths in a state space graph. The behavior of the algorithm is illustrated on a small example. Finally, an axiomatic justification of the refinement of robustness by an OWA criterion is provided. 1
MetaLevel Control for DecisionTheoretic Planners
, 1996
"... MetaLevel Control Agents plan in order to improve their performance, but planning takes time and other resources that can degrade performance. To plan effectively, an agent needs to be able to create high quality plans efficiently. Artificial intelligence planning techniques provide methods for gen ..."
Abstract

Cited by 11 (1 self)
 Add to MetaCart
MetaLevel Control Agents plan in order to improve their performance, but planning takes time and other resources that can degrade performance. To plan effectively, an agent needs to be able to create high quality plans efficiently. Artificial intelligence planning techniques provide methods for generating plans, whereas decision theory offers expected utility as a measure for assessing plan quality, taking the value of each outcome and its likelihood into account. The benefits of combining artificial intelligence planning techniques and decision theory have long been recognized. However, these benefits will remain unrealized if the resulting decisiontheoretic planners cannot generate plans with high expected utility in a timely fashion. In this dissertation, we address the metalevel control problem of allocating computation to make decisiontheoretic planning efficient and effective. For efficiency, decisiontheoretic planners iteratively approximate the complete solution to a decision problem: planners generate partially elaborated, abstract plans; only promising plans are further refined, and execution may begin before a plan with the highest expected
Improved Results for Route Planning in Stochastic Transportation Networks
 In Proc. of Symposium of Discrete Algorithms
, 2001
"... In the bus network problem, the goal is to generate a plan for getting from point X to point Y within a city using buses in the smallest expected time. Because bus arrival times are not determined by a xed schedule but instead may be random, the problem requires more than standard shortest path t ..."
Abstract

Cited by 11 (1 self)
 Add to MetaCart
(Show Context)
In the bus network problem, the goal is to generate a plan for getting from point X to point Y within a city using buses in the smallest expected time. Because bus arrival times are not determined by a xed schedule but instead may be random, the problem requires more than standard shortest path techniques. In recent work, Datar and Ranade provide algorithms in the case where bus arrivals are assumed to be independent and exponentially distributed. We oer solutions to two important generalizations of the problem, answering open questions posed by Datar and Ranade. First, we provide a polynomial time algorithm for a much wider class of arrival distributions, namely those with increasing failure rate. This class includes not only exponential distributions but also uniform, normal, and gamma distributions. Second, in the case where bus arrival times are independent and geometric discrete random variables, we provide an algorithm for transportation networks of buses and trains...
Optimal Factory Scheduling using Stochastic Dominance A*
 In Proceedings of the Twelfth Conference on Uncertainty in Artificial Intelligence
, 1996
"... We examine a standard factory scheduling problem with stochastic processing and setup times, minimizing the expectation of the weighted number of tardy jobs. Because the costs of operators in the schedule are stochastic and sequence dependent, standard dynamic programming algorithms such as A* may f ..."
Abstract

Cited by 10 (2 self)
 Add to MetaCart
We examine a standard factory scheduling problem with stochastic processing and setup times, minimizing the expectation of the weighted number of tardy jobs. Because the costs of operators in the schedule are stochastic and sequence dependent, standard dynamic programming algorithms such as A* may fail to find the optimal schedule. The SDA* (Stochastic Dominance A*) algorithm remedies this difficulty by relaxing the pruning condition. We present an improved statespace search formulation for these problems and discuss the conditions under which stochastic scheduling problems can be solved optimally using SDA*. In empirical testing on randomly generated problems, we found that in 70%, the expected cost of the optimal stochastic solution is lower than that of the solution derived using a deterministic approximation, with comparable search effort. 1 INTRODUCTION Generating production schedules for manufacturing facilities is a problem of great theoretical and practical importance. The Op...
Search for Choquetoptimal paths under uncertainty
 in UAI’07, 2007
"... Choquet expected utility (CEU) is one of the most sophisticated decision criteria used in decision theory under uncertainty. It provides a generalisation of expected utility enhancing both descriptive and prescriptive possibilities. In this paper, we investigate the use of CEU for pathplanning unde ..."
Abstract

Cited by 8 (3 self)
 Add to MetaCart
(Show Context)
Choquet expected utility (CEU) is one of the most sophisticated decision criteria used in decision theory under uncertainty. It provides a generalisation of expected utility enhancing both descriptive and prescriptive possibilities. In this paper, we investigate the use of CEU for pathplanning under uncertainty with a special focus on robust solutions. We first recall the main features of the CEU model and introduce some examples showing its descriptive potential. Then we focus on the search for Choquetoptimal paths in multivalued implicit graphs where costs depend on different scenarios. After discussing complexity issues, we propose two different heuristic search algorithms to solve the problem. Finally, numerical experiments are reported, showing the practical efficiency of the proposed algorithms. 1
Using StochasticDominance Relationships for Bounding Travel Times in Stochastic Networks
, 1999
"... We consider stochastic networks' in which link travel times are dependent, discrete random variables. We present methods' for computing bounds' on path travel times using stochastic dominance relationships among link travel times, and discuss techniques for controlling tightness of th ..."
Abstract

Cited by 7 (5 self)
 Add to MetaCart
We consider stochastic networks' in which link travel times are dependent, discrete random variables. We present methods' for computing bounds' on path travel times using stochastic dominance relationships among link travel times, and discuss techniques for controlling tightness of the bounds'. We apply these methods' to shortestpath problems, show that the proposed algorithm can provide bounds' on the recommended path, and elaborate on extensions of the algorithm for demonstrating the anytime property.