Temporal Reasoning with Preferences and Uncertainty

Temporal Constraint Satisfaction Problems allow for reasoning with events happening over time. Their expressiveness has been extended independently in two directions: to account for uncontrollable events, and, more recently, to account for soft temporal preferences. The motivation for both extensions is from real-life temporal problems; and indeed such problems may well necessitate both preferences and uncertainty. This paper proposes the study of temporal problems with both preferences and uncertainty, and puts forward some methods for their resolution. 1 Motivation and Background Research on temporal reasoning, once exposed to the difficulties of real-life problems, can be found lacking both expressiveness and flexibility. Planning and scheduling for satellite observations, for instance, involves not only quantitative temporal constraints between events and qualitative temporal ordering of events, but also soft temporal preferences and contingent events over which the agent has no control. For example, on one hand, slewing and scanning activities should not overlap, but may if necessary. On the other hand, the duration of failure recovery procedures is not under the direct control of the satellite executive. To address the lack of expressiveness, preferences can be added to the framework; to address the lack of flexibility to contingency, handling of uncertainty can be added. Some real-world problems, however, have need for both. It is this requirement that motivates us. In a temporal constraint problem, as defined in Dechter et al [1991], variables denote timepoints and constraints represent the possible temporal relations between them. Such constraints are quantitative, describing restrictions on either durations or distances of events, in terms of intervals over the timeline. In general such problems are NP-complete. However, if each temporal constraint has just one interval — hence the constraints have form where denote the timepoints — then we have a Simple Temporal Problem (STP) that can be solved in polynomial time. 'Partially supported by the EPSRC under grant GR/N64373/01. f Partially funded by the Italian MIUR projects NAPOLI and COVER, and the ASI (Italian Space Agency) project ARISCOM. To address the lack of expressiveness in standard STPs, the Simple Temporal Problems with Preferences (STPP) framework merge STPs with semiring-based soft constraints [Rossi et al, 2002]. Soft temporal constraints are specified by means of a preference junction on the constraint interval, where A is a set of preference values. The set A is part of a semiring. In general, STPPs are NP-complete. However, if the preference functions are semi-convex (that is, they have at most one peak), the constraints are combined via an idempotent operation (like max or min), and the preference values are totally ordered, then finding an optimal solution of a STPP is a polynomial problem. Two solvers for STPPs with the features above are presented by Rossi et al. [2002]. The first, Path-solver, enforces path consistency in the constraint network, then takes the subinterval on each constraint corresponding to the best preference level. This gives a standard STP, which is then solved for the first solution by backtrack-free search. The second solver, Chop-solver, is less general but more efficient. It finds the maximum level y at which the preference functions can be 'chopped', i.e. reduced to the set This set is a simple interval for each I. Hence we obtain a standard STP, STPy. By binary search, the solver finds the maximal y for which STPy is consistent. The solutions of this STP are the optimal solutions of the original STPP. To address the lack of flexibility in execution of standard STPs, Vidal and Fargier [1999] introduced Simple Temporal Problems under Uncertainty (STPUs). Here, again as in a STP, the activities have durations given by intervals. The timepoints, however, fall into two classes: requirement and contingent. The former, as in a STP, are decided by the agent, but the latter are decided by 'nature': the agent has no control over when the activity will end; he observes rather than executes. The only information known prior to observation is that nature wil l respect the interval on the duration. Controllability of a STPU is the analogue of consistency of a STP Controllable implies the agent has a means to execute the timepoints under his control, subject to all constraints. Three notions are proposed. A STPU is strongly controllable if there is a fixed execution strategy that works in all realisations (that is, an observation of all contingent timepoints). Checking strong controllability is in P [Vidal and Fargier, 1999]. A STPU is dynamically controllable if there is a online execution strategy that depends only on observed timepoints in the POSTER PAPERS 1385 N. Yorke-Smith IC-Parc Imperial College London London, U.K. nys@icparc . i c .ac .uk* K. B. Venable and F. Rossi Dept. of Mathematics University of Padova Padova, Italy { kvenab le , f ross i }@math .un ipd . i t * past and that can always be extended to a complete schedule whatever may happen in the future. Checking dynamic controllability is also in P [Morris et al, 2001]. A STPU is weakly controllable if there exists at least one execution strategy for every realisation. Checking weak controllability is co-NP [Vidal and Fargier, 1999]. The three notions are ordered by their strength; strong dynamic weak. In this paper we introduce a framework which allows one to handle both preferences and uncertainty in temporal problems. This is done by just merging the two pre-existing models of STPPs and STPUs. We also adopt the notion of controllability, to be used instead of consistency because of the presence of uncertainty, and we adapt it to handle preferences. We then investigate the complexity of checking such notions of controllability, and we develop algorithms to perform such checks. The main results are for strong and weak controllability, which we show to maintain the same complexity as in STPUs under the same hypothesis that make STPPs polynomia l ^ solvable. As future work, we plan to study dynamic controllability, and to implement our algorithms to compare their performance experimentally. 2 Simple Temporal Problems with Preferences and Uncertainty (STPPU) A STPPU is a STPP with timepoints partitioned into two classes, requirement and contingent, just as in a STPU. Since some timepoints are not controllable by the agent, the notion of consistency of a STP(P) is replaced by controllability, just as in a STPU, with the three notions suitably extended. While all three notions of controllability used for STPUs can be adapted to STPPUs, for this short paper we wil l focus on just weak and strong controllability. A STPPU is optimally strongly controllable if there is a fixed execution strategy that works in all realisations and is optimal in each of them. Rather than optimality, one could be interested in just reaching a certain quality level: a STPPU is a-strongly controllable, where a is a preference level, if there is a fixed execution strategy that works in all realisations and is above a in each of them. A STPPU is optimally (resp. a-) weakly controllable if there exists at least one optimal (resp., above a) execution strategy for every realisation. 3 Checking Optimal and a Strong