This paper presents a manifesto that articulates the conceptual foundations of Agentic Business Process Management (APM), an extension of Business Process Management (BPM) for governing autonomous agents executing processes in organizations. From a management perspective, APM represents a paradigm shift from the traditional view on business processes. This shift is driven by the realization of process awareness by agent-oriented abstractions: software and human agents act as primary functional entities that perceive, reason, and act within explicit process frames. Thus, APM moves away from automation-oriented BPM towards systems in which autonomy is constrained, aligned, and made operational through process aware agents. We introduce the core abstractions and architectural elements required to realize APM systems and elaborate on four key capabilities that agents in APM systems must support: framed autonomy , explainability , conversational actionability , and self-modification . These capabilities jointly ensure that agents’ goals are aligned with organizational goals and that agents behave in a framed yet proactive manner in pursuing those goals. We discuss the extent to which the capabilities can be realized and identify research challenges whose resolution requires further advances in BPM, AI, and multi-agent systems. The manifesto thus serves as a roadmap for bridging these communities and for guiding the development of APM systems in practice.

The Business Process Modeling Notation (BPMN) is the de facto standard for business process modeling. While widely adopted for its intuitive graphical notation, its execution semantics described in natural language lacks a commonly agreed formal foundation, leading to variability in execution across different BPM systems (BPMSs) and increasing the risk of creating models with semantic errors costly to correct at runtime. Although many formalisms have been used to model portions of BPMN, their reasoning capabilities are mostly restricted to control-flow, making them unsuitable for semantic analysis where data and global exception handling play a central role in execution. To address this, we propose a formalization from BPMN to ConGolog, a logical concurrent processes language based on the Situation Calculus, for representing and reasoning about dynamic domains. A major innovation is using ConGolog to rigorously capture the semantics of BPMN global exceptions. Our framework supports advanced reasoning, allowing for semantic analysis of BPMN models before execution to predict runtime errors within a safe simulation setting, while laying the foundation for reasoning layers in next-generation AI-augmented BPMSs. We validate the approach through a prototype and comprehensive evaluation, demonstrating the computational feasibility of the translation and the semantic correctness of reasoning tasks.

We analyse the computational complexity of two types of partial satisfaction planning problems, namely, over-subscription planning and net-benefit planning. Partial satisfaction planning seeks to maximise the rewards, or net-benefit between the rewards and costs, that can be obtained by achieving a set of (candidate) goals, generally under a cost budget. In this paper, we assume that the cost budget is polynomially bounded with respect to the number of atoms. In contrast to previous works on complexity analysis, we view partial satisfaction planning as optimisation tasks instead of decision ones. Accordingly, we classify over-subscription and net-benefit planning problems in classes within the optimisation complexity framework. Let L be the size of the partial satisfaction planning instance, we first show that both optimisation problems are OptP[LO(1)]-complete. We then prove that while over-subscription planning becomes OptP[O(log L)]-complete under uniform goal rewards and uniform action costs, the complexity of net-benefit planning remains unchanged. Our results show that the relative values of rewards and costs may impact the computational complexity of partial satisfaction planning, and that an optimisation perspective can reveal differences that may otherwise be lost in their decision formulations. In addition, an optimisation complexity analysis can provide insights on approximation properties of these problems.

This volume contains the papers accepted for presentation at ICAPS 2025, the Thirty-Fifth International Conference on Automated Planning and Scheduling, to be held in Melbourne, Australia, November 9-14, 2025. The annual ICAPS conference series was formed in 2003 through the merger of two pre-existing biennial conferences, the International Conference on Artificial Intelligence Planning and Scheduling (AIPS) and the European Conference on Planning (ECP). ICAPS continues the traditional high standards of AIPS and ECP as an archival forum for new research in the field of automated planning and scheduling. ICAPS 2025 was co-located with two other events: The International Conference on the Integration of Constraint Programming, Artificial Intelligence, and Operations Research (CPAIOR), and The International Conference on the Principles of Knowledge Representation and Reasoning (KR). Existing research into methods and representations for Automated Planning and Scheduling has drawn heavily from the research conducted by these communities. We believe that the co-location of these conferences with ICAPS can only boost these beneficial relationships.The frontmatter contains a Preface and lists both the ICAPS 2025 Organising Committee and the ICAPS 2025 Program Committee.

We propose an approach for solving Agent Planning Programs (APP) based on a reduction to (strong-cyclic) Fully Observable Non-Deterministic (FOND) planning. APPs represent a middle-ground between automated planning and agent-oriented programming, in which the space of possible agent behavior is "programmed" as a network of declarative goals wrt an underlying planning domain. Each transition in an APP represents a local planning problem that may need to be addressed by the agent executing the APP. APPs allow the specification of continuous goal-driven behavior in which the "next" goal is externally chosen, thus going beyond one-shot planning. Two methods have been proposed for solving APPs: a principled but inefficient LTL reactive synthesis technique; and a more efficient but arguably ad-hoc approach that relies on multiple "local" classical planning and meta-level backtracking. We demonstrate how APPs can be solved in a principled manner by developing an elegant reduction to non-deterministic planning under fairness assumptions, and show experimentally with existing FOND solvers its practical value. We also provide a new solution concept that is simpler and closer to mainstream planning than the existing one.

Process mining (PM)-based goal recognition (GR) techniques, which infer goals or targets based on sequences of observed actions, have shown efficacy in real-world engineering applications. This study explores the applicability of PM-based GR in identifying target poses for users employing powered transhumeral prosthetics. These prosthetics are designed to restore missing anatomical segments below the shoulder, including the hand. In this article, we aim to apply the GR techniques to identify the intended movements of users, enabling the motors on the powered transhumeral prosthesis to execute the desired motions precisely. In this way, a powered transhumeral prosthesis can assist individuals with disabilities in completing movement tasks. PM-based GR techniques were initially designed to infer goals from sequences of observed actions, where discrete event names represent actions. However, the electromyography electrodes and kinematic sensors on powered transhumeral prosthetic devices register sequences of continuous, real-valued data measurements. Therefore, we rely on methods to transform sensor data into discrete events and integrate these methods with the PM-based GR system to develop target pose recognition approaches. Two data transformation approaches are introduced. The first approach relies on the clustering of data measurements collected before the target pose is reached (the clustering approach). The second approach uses the time series of measurements collected while the dynamic user movement to perform linear discriminant analysis (LDA) classification and identify discrete events (the dynamic LDA approach). These methods are evaluated through offline and human-in-the-loop (online) experiments and compared with established techniques, such as static LDA, an LDA classification based on data collected at static target poses, and GR approaches based on neural networks. Real-time human-in-the-loop experiments further validate the effectiveness of the proposed methods, demonstrating that PM-based GR using the dynamic LDA classifier achieves superior F 1 score and balanced accuracy compared to state-of-the-art techniques.

Goal Recognition (GR) is a research problem that studies ways to infer the goal of an intelligent agent based on its observed behavior and knowledge of the environment in which the agent operates. A common assumption of GR is that the environment is static. However, in many real-world scenarios, for example, recognizing customers’ preferences, it is necessary to recognize the goals of multiple agents or multiple goals of a single agent over an extended period. Therefore, it is reasonable to expect the environment to change throughout a series of goal recognition tasks. This paper presents three process mining-based solutions to the problem of adaptive GR in a changing environment implemented as different control strategies of a system for solving standard GR problems. As a standard GR system that gets controlled, we use the system grounded in process mining techniques, as it can adjust its internal GR mechanisms based on data collected while observing the operating agents. We evaluated our control strategies over synthetic and real-world datasets. The synthetic datasets were generated using the extended version of the Goal Recognition Amidst Changing Environments (GRACE) tool. The datasets account for different types of changes and drifts in the environment. The evaluation results demonstrate a trade-off between the GR performance over time and the effort invested in adaptations of the GR mechanisms of the system, showing that few well-planned adaptations can lead to a consistently high GR performance.

We present an approach to non-deterministic planning under full observability via Answer Set Programming. The technique can synthesise compact policies, handle both fair and unfair actions simultaneously, and readily accommodate control knowledge and procedural domain constraints. We show that whereas compact controllers may yield sub-optimal behavior under a naive executor, optimality can be recovered under a smarter executor. The developed planner is succinct, elegant, and directly implementable, thus providing higher confidence of its correctness and ease of elaboration. Experimental results show that its performance is competitive.

A transhumeral prosthesis restores missing anatomical segments below the shoulder, including the hand. Active prostheses utilize real-valued, continuous sensor data to recognize patient target poses, or goals, and proactively move the artificial limb. Previous studies have examined how well the data collected in stationary poses, without considering the time steps, can help discriminate the goals. In this case study paper, we focus on using time series data from surface electromyography electrodes and kinematic sensors to sequentially recognize patients' goals. Our approach involves transforming the data into discrete events and training an existing process mining-based goal recognition system. Results from data collected in a virtual reality setting with ten subjects demonstrate the effectiveness of our proposed goal recognition approach, which achieves significantly better precision and recall than the state-of-the-art machine learning techniques and is less confident when wrong, which is beneficial when approximating smoother movements of prostheses.

The problem of goal recognition requests to automatically infer an accurate probability distribution over possible goals an autonomous agent is attempting to achieve in the environment. The state-of-the-art approaches for goal recognition operate under full knowledge of the environment and possible operations the agent can take. This knowledge, however, is often not available in real-world applications. Given historical observations of the agents' behaviors in the environment, we learn skill models that capture how the agents achieved the goals in the past. Next, given fresh observations of an agent, we infer their goals by diagnosing deviations between the observations and all the available skill models. We present a framework that serves as an outline for implementing such data-driven goal recognition systems and its instance system implemented using process mining techniques. The evaluations we conducted using our publicly available implementation confirm that the approach is well-defined, i.e., all system parameters impact its performance, has high accuracy over a wide range of synthetic and real-world domains, which is comparable with the more knowledge-demanding state-of-the-art approaches, and operates fast.

Manufacturing is transitioning from a mass production model to a service model in which facilities `bid' for previously unseen products. To decide whether to bid for a previously unseen product, a facility must be able to synthesize, on the fly, a process plan controller that delegates abstract manufacturing tasks in a supplied process recipe to the available manufacturing resources. First-order representations of the state are commonly considered in reasoning about action in AI. Here we show that we can leverage the wide literature on the Situation Calculus automatically synthesize such controllers. We identify two important decidable cases---finite domains and bounded action theories---for which we provide practical synthesis techniques.

We consider the problem of reaching a propositional goal condition in fully-observable nondeterministic (FOND) planning under a general class of fairness assumptions that are given explicitly. The fairness assumptions are of the form A/B and say that state trajectories that contain infinite occurrences of an action a from A in a state s and finite occurrence of actions from B, must also contain infinite occurrences of action a in s followed by each one of its possible outcomes. The infinite trajectories that violate this condition are deemed as unfair, and the solutions are policies for which all the fair trajectories reach a goal state. We show that strong and strong-cyclic FOND planning, as well as QNP planning, a planning model introduced recently for generalized planning, are all special cases of FOND planning with fairness assumptions of this form which can also be combined. FOND+ planning, as this form of planning is called, combines the syntax of FOND planning with some of the versatility of LTL for expressing fairness constraints. A sound and complete FOND+ planner is implemented by reducing FOND+ planning to answer set programs, and its performance is evaluated in comparison with FOND and QNP planners, and LTL synthesis tools. Two other FOND+ planners are introduced as well which are more scalable but are not complete.

Many cognitive problems people face have been shown to be computationally intractable. However, tractability is typically defined in terms of asymptotic worst-case behaviour of instances. One approach for studying typical cases of NP-complete problems is based on random instances. It has been shown that random instances of many NP-complete problems exhibit a phase transition in solvability and that hard instances tend to occur in this phase transition. Here, we characterise a phase transition in solvability for random instances of the 0-1 knapsack problem in terms of two simple instance properties. Subsequently, we show that compute time of algorithms peaks in the phase transition. Remarkably, the phase transition likewise predicts where people spend the most effort. Nevertheless, their performance decreases. This suggests that instances that are difficult for electronic computers are recognized as such by people, but the increased effort does not compensate for hardness. Given the ubiquity of the knapsack problem in every-day life, a better characterisation of the properties that make instances hard will help understand commonalities and differences in computation between human and digital computers, and to improve both decision environments (contracts, regulation) as well as human-computer interfaces.

This work investigates the problem of optimising a partial-order plan’s (POP) flexibility through the simultaneous transformation of its action ordering and variable binding constraints. While the former has been extensively studied through the notions of deordering and reordering, the latter has received much less attention. We show that a plan’s variable bindings are often related to resource usage and their reinstantiation can yield more flexible plans. To do so, we extend existing POP optimality criteria to support variable reinstantiation, and prove that checking if a plan can be optimised further is NP-complete. We also propose a MaxSAT-based technique for increasing plan flexibility and provide a thorough experimental evaluation that suggests that there are benefits in action reinstantiation.

Goal recognition is the problem of determining an agent's intent by observing her behaviour. Contemporary solutions for general task-planning relate the probability of a goal to the cost of reaching it. We adapt this approach to goal recognition in the strict context of path-planning. We show (1) that a simpler formula provides an identical result to current state-of-the-art in less than half the time under all but one set of conditions. Further, we prove (2) that the probability distribution based on this technique is independent of an agent's past behaviour and present a revised formula that achieves goal recognition by reference to the agent's starting point and current location only. Building on this, we demonstrate (3) that a Radius of Maximum Probability (i.e., the distance from a goal within which that goal is guaranteed to be the most probable) can be calculated from relative cost-distances between the candidate goals and a start location, without needing to calculate any actual probabilities. In this extended version of earlier work, we generalise our framework to the continuous domain and discuss our results, including the conditions under which our findings can be generalised back to goal recognition in general task-planning.

In this letter, we report on the use of the UC-Berkeley Pacman Capture the Flag competition as a major assessment component in two large artificial intelligence (AI) courses. Teams of students are to integrate AI techniques to control the behavior of agents in an adversarial version of the well-knownPacmangame. We discuss how the contest was designed, set up, and used for grading, and reflect on some of the outcomes achieved, including student perception. While the contest is very challenging (and, at times, frustrating) for students, it was successful in keeping them engaged. More importantly, a significant majority of students reported enjoying and learning from it. A powerful software tool to automate the competition for large number of teams is made freely available.

"Plan recognition as planning" uses an off-the-shelf planner to perform goal recognition. In this paper, we apply the technique to path-planning. We show that a simpler formula provides an identical result in all but one set of conditions and, further, that identical ranking of goals by probability can be achieved without using any observations other than the agent's start location and where she is "now".

We contribute to recent efforts in relating two approaches to automatic synthesis, namely, automated planning and discrete reactive synthesis. First, we develop a declarative characterization of the standard "fairness" assumption on environments in non-deterministic planning, and show that strong-cyclic plans are correct solution concepts for fair environments. This complements, and arguably completes, the existing foundational work on non-deterministic planning, which focuses on characterizing (and computing) plans enjoying special "structural" properties, namely loopy but closed policy structures. Second, we provide an encoding suitable for reactive synthesis that avoids the naive exponential state space blowup. To do so, special care has to be taken to specify the fairness assumption on the environment in a succinct manner.

We propose an approach to reason about agent planning programs with global information. Agent planning programs can be understood as a network of planning tasks, accommodating long-term goals, non-terminating behaviors, and interactive execution. We provide a technique that relies on reasoning about ``global" dead-ends and that can be incorporated to any planning-based approach to agent planning problems. In doing so, we also introduce the notion of online execution of such planning structures. We provide experimental evidence suggesting the technique yields significant benefits.

Deceptive path-planning involves finding a path such that the probability of an observer identifying the path's final destination - before it has been reached - is minimised. This paper formalises deception as it applies to path-planning and introduces the notion of a last deceptive point (LDP) which, when measured in terms of 'path completion', can be used to rank paths by their potential to deceive. Building on recent developments in probabilistic goal-recognition, we propose a formula to calculate an optimal LDP and present strategies for the generation of deceptive paths by both simulation ('showing the false') and dissimulation ('hiding the real').

In this work we present a novel approach to check the consistency of agent designs (prior to any implementation) with respect to the requirements specifications via automated planning. This checking is essentially a search problem which makes planning technology an appropriate solution. We focus our work on BDI agent systems and the Prometheus design methodology in order to directly compare our approach to previous work. Our experiments in more than 16K random instances prove that the approach is more effective than previous ones proposed: it achieves higher coverage, lower run-time, and importantly, can handle loops in the agent detailed design and unbounded subgoal reasoning.

In this paper, a probabilistic space-time representation of complex traffic scenarios is predicted using machine learning algorithms. Such a representation is significant for all active vehicle safety applications especially when performing dynamic maneuvers in a complex traffic scenario. As a first step, a hierarchical situation classifier is used to distinguish the different types of traffic scenarios. This classifier is responsible for identifying the type of the road infrastructure and the safety-relevant traffic participants of the driving environment. With each class representing similar traffic scenarios, a set of Random Forests (RFs) is individually trained to predict the probabilistic space-time representation, which depicts the future behavior of traffic participants. This representation is termed as a Predicted-Occupancy Grid (POG). The input to the RFs is an Augmented Occupancy Grid (AOG). In order to increase the learning accuracy of the RFs and to perform better predictions, the AOG is reduced to low-dimensional features using a Stacked Denoising Autoencoder (SDA). The excellent performance of the proposed machine learning approach consisting of SDAs and RFs is demonstrated in simulations and in experiments with real vehicles. An application of POGs to estimate the criticality of traffic scenarios and to determine safe trajectories is also presented.

The increasing application of process-oriented approaches in new challenging dynamic domains beyond business computing (e.g., healthcare, emergency management, factories of the future, home automation, etc.) has led to reconsider the level of flexibility and support required to manage complex knowledge-intensive processes in such domains. A knowledge-intensive process is influenced by user decision making and coupled with contextual data and knowledge production, and involves performing complex tasks in the “physical” real world to achieve a common goal. The physical world, however, is not entirely predictable, and knowledge-intensive processes must be robust to unexpected conditions and adaptable to unanticipated exceptions, recognizing that in real-world environments it is not adequate to assume that all possible recovery activities can be predefined for dealing with the exceptions that can ensue. To tackle this issue, in this paper we present SmartPM, a model and a prototype Process Management System featuring a set of techniques providing support for automated adaptation of knowledge-intensive processes at runtime. Such techniques are able to automatically adapt process instances when unanticipated exceptions occur, without explicitly defining policies to recover from exceptions and without the intervention of domain experts at runtime, aiming at reducing error-prone and costly manual ad-hoc changes, and thus at relieving users from complex adaptations tasks. To accomplish this, we make use of well-established techniques and frameworks from Artificial Intelligence, such as situation calculus, IndiGolog and classical planning. The approach, which is backed by a formal model, has been implemented and validated with a case study based on real knowledge-intensive processes coming from an emergency management domain.

We address verification of high-level programs over situation calculus action theories that have an infinite object domain, but bounded fluent extensions in each situation. We show that verification of mu-calculus temporal properties against ConGolog programs over such bounded theories is decidable in general. To do this, we reformulate the transition semantics of ConGolog to keep the bindings of “pick variables” into a separate variable environment whose size is naturally bounded by the number of variables. We also show that for situation-determined ConGolog programs, we can compile away the program into the action theory itself without loss of generality. This can also be done for arbitrary programs, but only to check certain properties, such as if a situation is the result of a program execution, not for mu-calculus verification.

Hierarchically structured agent plans are important for efficient planning and acting, and they also serve (among other things) to produce “richer” classical plans, composed not just of a sequence of primitive actions, but also “abstract” ones representing the supplied hierarchies. A crucial step for this and other approaches is deriving precondition and effect “summaries” from a given plan hierarchy. This paper provides mechanisms to do this for more pragmatic and conventional hierarchies than in the past. To this end, we formally define the notion of a precondition and an effect for a hierarchical plan; we present data structures and algorithms for automatically deriving this information; and we analyse the properties of the presented algorithms. We conclude the paper by detailing how our algorithms may be used together with a classical planner in order to obtain abstract plans.

In this paper, we review a series of agent behavior synthesis problems under full observability and nondeterminism (partial controllability), ranging from conditional planning, to recently introduced agent planning programs, and to sophisticated forms of agent behavior compositions, and show that all of them can be solved by model checking two-player game structures. These structures are akin to transition systems/Kripke structures, usually adopted in model checking, except that they distinguish (and hence allow to separately quantify) between the actions/moves of two antagonistic players. We show that using them we can implement solvers for several agent behavior synthesis problems.

When it comes to building controllers for robots or agents, high level programming languages like Golog and ConGolog offer a useful compromise between planning-based approaches and low-level robot programming. However, two serious problems typically emerge in practical implementations of these languages: how to evaluate test in a program efficiently enough in an open-world setting, and how to make appropiate nondeterministic choices while avoiding full lookahead. Recent proposals in the literature suggest that one could tackle the first problem by exploiting sensing information, and tackle the second by specifying the amount of lookahead allowed explicitly in the program. In this paper, we combine these two ideas and demonstrate their power by presenting an interpreter, written in Prolog, for a variant of Golog that is suitable for efficiently operating in open-world setting by exploiting sensing and bounded lookahead.