Call for Papers: ICAPS 2024

The 34th International Conference on Automated Planning and Scheduling (ICAPS 2024) will take place in Banff, Alberta (Canada), June 1-6, 2024. ICAPS 2024 is a part of the ICAPS conference series, the premier forum for exchanging news and research results on the theory and applications of intelligent planning and scheduling technology.

Traditionally, ICAPS has had both a main track and a set of specialized tracks (e.g., Robotics, Applications, etc.) that authors could submit to, all mutually exclusive. This year, we will have a single main track, with former track themes replaced by a set of optional primary keywords that distinguish papers for differential review within the main track. These optional keywords will be in addition to the standard keywords we use every year. The purpose of these primary keywords is to enable submissions to self-identify research themes, support the identification of suitable reviewers, and potentially aid in the grouping of thematically related accepted papers in the conference program. Authors (resp. reviewers) will be able to optionally select one or more primary keywords to reflect best the thematic contributions of the paper (resp. reviewer expertise). Many authors will not avail themselves of these optional primary keywords.

Primary Keywords

  1. Applications
  2. Learning
  3. Robotics
  4. Theory
  5. Human-aware Planning and Scheduling
  6. Temporal Planning
  7. Multi-Agent Planning
  8. Knowledge Representation/Engineering

A short description of each primary keyword and expectations of papers/reviewers associated with them is provided below. Additionally, the ICAPS program will include a journal track, demonstrations, workshops, and tutorials, each having separate submission and notification dates, to be announced separately.

The ICAPS 2024 program committee invites paper submissions related to automated planning and scheduling. Relevant contributions include but are not limited to:

  • Theoretical foundations of planning and scheduling
  • Classical planning techniques and analysis
  • Knowledge representation for planning and scheduling
  • Planning with time and resources
  • Plan and schedule execution, monitoring, and repair
  • Continual, on-line, and real-time planning and scheduling
  • Plan recognition, plan management, and goal reasoning
  • Uncertainty and stochasticity in planning and scheduling
  • Partially observable and unobservable domains
  • Conformant, contingent, and adversarial planning
  • Multi-agent and distributed planning
  • Planning and scheduling with continuous state and action spaces
  • Planning and scheduling with mixed continuous and discrete states/actions/decisions
  • Robot planning and scheduling
  • Knowledge acquisition and engineering for planning and scheduling
  • Applications and case studies of planning and scheduling techniques
  • Learning for planning and scheduling
  • Human-aware planning and scheduling

Author Guidelines

Authors may submit technical papers or position papers. Technical papers are expected to report on substantial original research, while position papers are expected to discuss ideas and concepts, for instance, via thoughtful critiques or bold new perspectives of the field, historical perspectives and analysis, or technical discussions of various implementation techniques. Both technical and position papers can be long (8 pages AAAI style plus up to one page of references) or short (4 pages plus up to one page of references). The type of paper must be indicated at submission time. All papers, regardless of length, will be reviewed against standard criteria such as relevance, originality, significance, clarity, and soundness and are expected to meet the same high standards set by ICAPS. Short papers may be of narrower scope, for example, by addressing a highly specific issue or proposing or evaluating a small, yet important, extension of previous work or new idea. Over-length papers will be rejected without review.

Authors making multiple submissions must ensure that each submission has significant, unique content. Papers submitted to ICAPS 2024 may not be submitted to other conferences or journals during the ICAPS 2024 review period, nor may they be already under review, accepted, or published in other conferences or journals.

We expect for ICAPS 2024 to be an in-person conference – one author of all accepted papers will be expected to present work in person unless there are exceptional circumstances that prevent this.

Use of ChatGPT or similar LLMs

ICAPS 2024 is adopting the AAAI policy regarding the use of ChatGPT and similar large-scale language models (LLMs). Papers that include text generated from an LLM, such as ChatGPT, are prohibited unless the produced text is presented as a part of the paper’s experimental analysis. Note that this policy does not prohibit authors from using LLMs for editing or polishing author-written text.

ICAPS 2024 furthermore follows AAAI policy in that any AI system, including Generative Models such as Chat-GPT, BARD, or DALL-E, does not satisfy the criteria for authorship of papers published by AAAI Press (ICAPS’ publisher) and, as such, also cannot be used as a citable source in papers published by AAAI press. Authors assume full responsibility for the content of submitted work, including checking for plagiarism and veracity of all text.

Best paper fast-track to top AI journals

Following the ICAPS tradition, in ICAPS 2024, we will give awards to the best paper and best student paper. These papers may also receive an invitation to submit an extended version to a top AI journal (e.g., the Journal of Artificial Intelligence Research (JAIR) or the Artificial Intelligence Journal (AIJ)).

Ethical/Societal Impact

It is optional for authors to include a statement of the potential broader impact of their work, including its ethical aspects and future societal consequences. This statement can be included in either the main body pages or the reference page. If such a statement is not included in the paper, but the reviewers deem that such a statement is necessary, then the authors will be asked to provide one during the author response period for review. If the paper is accepted, the statement provided will need to be incorporated into the camera-ready version.

Submission Instructions

All submissions will be made electronically via OpenReview.

Details on what the review form will look like will be publicly released soon.

Submitted papers should be anonymous for double-blind reviewing, adhere to the page limits of the relevant submission type (long or short), and follow the author kit instructions for formatting. A custom author kit for ICAPS 2024, based on the AAAI 2024 template can be found here:

In addition to the submitted PDF paper, authors can submit supplementary material (videos, technical proofs, additional experimental results) for their paper. Please make sure that the supporting material is also anonymized. Papers should be self-contained; reviewers are encouraged, but not obligated, to consider supporting material in their decisions.

Important Dates

We have arranged our schedule to allow for the revision and resubmission of papers rejected from AAAI 2024.

  • December 7, 2023 - Abstracts due (electronic submission)
  • December 13, 2023 - Full papers due (electronic submission, PDF)
  • January 24-27, 2024 - Author feedback period
  • February 12, 2024 - Notification of acceptance or rejection

The reference timezone for all deadlines is UTC-12. That is, as long as there is still some place anywhere in the world where the deadline has not yet passed, you are on time!

ICAPS 2023 Program Chairs

  • Sara Bernardini, Royal Holloway University of London, UK
  • Christian Muise, Queen’s University, Canada

For inquiries contact:


What will be made public from the review process?

The anonymized reviews, and final rebuttals from authors, will be made public on just those papers accepted. This will happen after the notification of acceptance/rejection. No discussion between reviewers and SPC will be made public, and reviewers/SPC will not be identified.

Aren’t you revealing too much?

We do not believe so. The reviews are already anonymized, and the rebuttals are written by the authors. We believe that the community will benefit from seeing the reviews and rebuttals, and that this will help improve the quality of the reviews in the long term.

Aren’t you revealing too little?

Historically, ICAPS has not made any part of the review process public. We believe that this is a good first step towards making the review process more transparent, and we will continue to evaluate what else can be made public in the future.

Aren’t the last two questions contradictory?

Yes, and we’ve been asked versions of both by severeal members of the community!

Can papers be revised while under review?

No, papers cannot be revised while under review. The author feedback period is for authors to respond to the reviews, not to revise their papers. There will be an opportunity to revise papers after the notification of acceptance/rejection before the camera-ready copy is due.

Is the rebuttal interactive/threaded?

No, the rebuttal is not interactive. Authors will be able to submit a single rebuttal to each review separately, and reviewers will be able to submit a single review. There will be no back-and-forth between authors and reviewers for this year.

What is the policy on uploading our submissions to arXiv?

Yes, following the AAAI policy, you are allowed to upload your submissions to arXiv during the review period.

Can rejected AAAI papers be submitted? The AAAI notification is after abstract submission.

We consider it a full submission only when you submit the full paper. So, yes, please do submit your abstract and withdraw the submission if it happens to be accepted to AAAI. If, for whatever reason, AAAI pushes back the notification beyond our paper submission deadline, we will reconsider that date.

Must I select a primary keyword?

No. We fully expect there to be several papers without a primary keyword selection. Only select primary keyword(s) if they make sense for your work.

What will the review form look like?

You can find a draft of the form questions reviewers will use [here].

Can we use more than a page of references if we use fewer than 8 pages of content?

Yes, you can use up to 9 pages total (5 pages for short papers), as long as the final page contains only references. This means that you may have 7.5 pages of content and 1.5 pages of references, for example.

Is there any exception to the limit of references?

We may grant specific exceptions to authors in situations such as a survey paper, where the broad list of references are a key part of the paper itself. Please contact the program chairs if you believe you have a special case.

How do we register conflicts?

This will come in two places. First, on your profile page, you can select conflicts to individuals (e.g., key advisors, collaborators, etc) – the link can be found [here]. Second, reviewers and SPC should be able to flag papers as conflicts during the bidding process. Some conflicts (e.g., papers where a reviewer is also an author) should be automatically registered.

Primary Keyword Descriptions


The goal of the “Applications” focus is to bring together planning and scheduling practitioners, industry stakeholders and academic communities for exchanging all aspects of the development, deployment, and evaluation of planning and scheduling systems for societal or industrial problems. We aim to bring scientific results from planning and scheduling to practice but also aim to identify scientific challenges by analyzing the limitations of current results to tackle real-world problems. We solicit participation from industry practitioners as well as from academics working on applications.


The aim of the “Learning” primary keyword is to identify papers that are at the intersection of machine or reinforcement learning on one hand and planning & scheduling on the other. Neurosymbolic approaches, data-driven domain or scheduling constraint learning, application of scheduling or planning to learning systems, etc., are all examples of this area. If relevant, authors are expected to use appropriate ML/RL methodology (e.g., train/validation/test splits, ablation studies, etc.), and reviewers will be expected to be well-versed in being able to assess the submitted works through this lens.


The “Robotics” primary keyword is to identify papers that revolve around the advancement of intelligent robots. Planning models and techniques are important for enabling autonomous, flexible, and interactive behaviours in robotic systems. A deep integration of these methods into robotic architectures can assist their effective deployment. Papers associated with this keyword would showcase the use of planning and scheduling technology in robotics applications. Examples of themes under this keyword are integrated task and motion planning, plan execution, failure detection and recovery, multi-robot planning, scheduling/coordination, execution, etc.


Papers that select the “Theory” keyword should contain novel contributions of a theoretical manner. Examples may include proving existing conjectures, introducing a novel proof technique, demonstrating novel complexity results, etc. Proving common properties of novel methods in otherwise core planning & scheduling techniques does not warrant the selection of “Theory” (e.g., showing the soundness/completeness of a new planning algorithm need not be flagged as “Theory”). Reviewers will be expected to be comfortable assessing the correctness and significance of theoretical contributions to the fields of planning & scheduling.

Human-Aware P&S

The “Human-Aware Planning and Scheduling” keyword refers to research contributions on planning, scheduling, and decision support systems that account for and fluidly interact with the humans in the loop. The keyword encompasses all works that look at tackling any challenges inherent to settings where a person needs to use, interact, collaborate or work along with planning and scheduling systems. Examples of themes under this keyword are plan and goal recognition, behavior prediction, mixed-initiative planning and scheduling systems, interpretability and explainability of AI planning/scheduling systems, trust, communication, and collaboration in human-AI teams, etc.

Temporal Planning

The “Temporal Planning” keyword refers to planning formulations where multiple actions can be taken simultaneously, their durations may vary, and actions and events may have complex interdependencies determining which combinations are possible. The keyword includes formulations where temporal constraints are defined in terms of restrictions on the ways actions in plans can be sequenced or overlap and where objective functions inspired by Operations Research are required (makespan, tardiness, etc.) to be optimized.

Multi-Agent Planning

The “Multi-Agent Planning” keyword refers to planning formulations where planning is performed by or for multiple agents. The keyword includes (but is not limited to) themes such as agents planning for a common goal, plan coordination/merging, plan negotiation, distributed planning and scheduling, real-time execution of multi-agent plans, multi-agent pathfinding, multi-agent epistemic planning, multi-agent systems, and applications of multi-agent systems.

Knowledge Representation/Engineering

The “Knowledge Representation/Engineering’’ keyword refers to representing agents’ knowledge explicitly in a declarative form. The keyword includes (but is not limited to) work on reformulation of models and domain knowledge, languages to represent knowledge, acquisition and refinement of models, translation of models between different languages, and injecting knowledge to support the search process (i.e., domain control knowledge).