Plans and Planning
          Representations
          Communicating
          Architectures and Environments
          Research Vehicles
          User Support and Methodologies
            I-X Research Meeting 26 Sept test
            Attendees
            Jussi Stader (JS)
            Gerhard Wickler (GW)
            Jessica Chen-Burger (JCB)
            Jeff Dalton (JD)
            Austin Tate (AT)
            Stephen Potter (SP)
            Purpose: Brainstorming potential research areas within the I-X programme
            Minutes: potential research areas
          Strategies for Research

            for instance using O-Plan goal structure - JD
            Deciding where to put effort in terms of search - AT
            need deconfliction - AT
            not quite state-of-the-art currently in I-Plan - GW
            plans to change availability change - AT
            Any intelligent agent must be prepared to explain their plan - AT
            Simulation of plan/world state -JD
            perhaps using plan matrix - JD
            ...planning for multiple agents - GW
            Requires user evaluation? - GW
            True mixed-initiative planning - AT
            ....also opportunistic execution - AT
            modelling of capability availability and changes to these via activity - AT
            ....and me! - JCB/AT
            People want this, rather than a planner as such - JD
            Multi-agent planning - GW
            What about chimps and dolphins? - JD
            Plan editing and repair - JD
            Distributed plan execution and repair - AT
            OPM Planner - AT
            Plan visualisation and explanation - JD
            Modular planning architecture - JD
            Plans for availability change - AT
            Planning and constraint algorithms - JD
            taking advantage of current state to achieve overall goals more efficiently - JS
            Action representations - GW
            representations that are expressive, with semantics and an ontology of actions - GW
            Opportunistic planning - JS
            need interaction resolution - AT
            Search control in I-Plan - GW
            relate to Strategic/Tactical/Operational levels - AT
            component analysis - modelling what individual agents plan for - AT
            The answer is , now what's the question? - AT
            making agents explicit in the plan (rather than treating them as resources) GW
            we already do this? - JD
            Plan comparison - JD
            ...planning by multiple agents - GW
            This interests me too - SP

            large amount of literature devoted to plan recognition - AT
            Ubiquitous computing - JCB
            Humans, robots, sensor grids, SW, teams - AT
            every telecoms/IP provider - AT
            Simulation of plan/world state -JD
            (very) large-scale deployments - AT
            Our interests are in assisting people in their task - AT
            "I'm less keen on this" - JD
            Assume that we know where/what they're doing... - JS
            Modular planning architecture - JD
            eg knowledge that person is in a particular place using a particular device affects interaction with them - JCB/JS
            The idea of sharing separate to the localisation of an individual is important - AT
            AT said "no" to UrAGE last week - AT
            maybe not good research focus? JD
            "database scale" really necessary? GW
            embedded in unchangable devices - AT
            modelling devices etc -JCB
            Should relate to David Wilkins work on plan servers - AT
            Persistent storage (plan database) - JD
            Are we interested in inferring this? - SP

            discovery of capabilities - AT
            the point about electronic means sounds restrictive - JCB
            SEBPC Programme - JS
            Steve Polyak studied QOC representations - AT
            trend analysis, histories, DM - AT
            Steve Polyak made a start on this with sorted FOL and properties of the space - AT
            INCA formal aspects - AT
            constraint-based representation - how do you work/reason with this? - AT
            a collection of agents, resources etc, working together toward some common goal, but which are not necessarily physically co-located, but which are primarily communicating and interacting using electronic means - SP
            separate panel linking to different agents, etc - JCB
            virtual organisation - SP
            Higher-level descriptions of organisational goals - JCB
            "what if" scenarios, options, simulations - AT
            I-World/I-Space ideas - AT
            use gIBIS issue types and study them - AT
            Action representations - GW
            Import/export from other representations - AT
            Modelling the change of entities (on the SW) - SP
            representations that are expressive, with semantics and an ontology of actions - GW
            This is what the PrintFair EU project is about - GW
            Authority and capability models and mechanisms for changing these - AT
            Business/Goal Modelling - JCB
            The answer is , now what's the question? - AT
            Objective/Task description - AT
            Data flow process models - JD
            Use - AT
            Organisational modelling - JCB
            Does this relate to Strategic/Tactical/Operational trichotomy - SP
            current state, previous state, potential future states - AT
            role modelling - JCB

            Compared to stand-alone editor - AT
            eg. like BidMan
            what would be the target representation for modelling? - AT
            richer domain modelling - JCB
            Modelling methodologies - SP
            Explore eg. Protege - AT
            Use whatever technique is most convenient/intuitive - JS
            ....but how is this structured/constrained? JS
            The answer is , now what's the question? - AT
            Support for I-X users - JS
            ontology merging - JS
            issue-based modelling - AT
            Is this domain object-oriented (or activity-oriented)? - AT
            support for the modelling process - AT
            Multi-perspective modelling - JS
            Generalising domain modeller - JS
            Import/export from other representations - AT
            intelligent domain modelling assistants - AT for JCB/JS
            so as not constrained by IP2 in particular - JS
            Process support that is information driven - JS

            link to BDI - AT
            ...for explanation - GW
            True mixed-initiative planning - AT
            Async vs. sync communications - AT
            Inter-agent communication modelling - SP
            speech acts - AT
            Terminology? - JD
            Dialogue between agents relating to tasks - AT
            eg. Dialogue planning for tutoring (Johanna Moore) - AT
            The context for the use of - AT
            Behaviour alters when you know people are on-line - AT

            Formulate, contribute to standards - AT
            Use of standards - AT
            Standards - AT
            Exercising - JS

            FireGrid
            Co-Opr
            AIBO
            OpenKnowledge/AKT e-Response
            IMPACs
            UrAGE
            Proforma:I-X link (CancerUK)
            PrintFair
            MERSI
            Work with MPAT
            JPRA PRETC
            Galileo
            smart cars
            entertainment companion robots
            environmental response systems
            Long-term
            Vehicles
            Proposals
            Current Projects
            Opportunities