Thomas Bolander, Hans van Ditmarsch, Jan van Eijck, Ramanujam September 2, 2015

Science: The workshop `To be announced!' was on the subject of epistemic protocols. In protocols (or planning), various aspects can be called epistemic: the condition for executing an action in a protocol may be knowledge or ignorance, such an action may

consist of imparting information (so, again, knowledge), actions may be partially observable (so that the agent does not know what really happened), and the goals of plans may be epistemic (the spy may not get to know the secret). Five workshop themes agency, concurrency, uncertainty, communication, strategy were investigated in discussion groups. There were 11 invited keynotes, who admirably focused on epistemic planning.

No tangible outcome of the workshop was planned, but over the week some became

clear. The discussion groups promoted collaborations between workshop participants. A

special issue of a journal is planned and was announced at the closing of the workshop.

The AI Journal or JAIR will be approached with a special issue proposal. The explicit goal of the workshop was to bring the `dynamic epistemic logic' and the `artificial intelligence' communities within epistemic planning closer. The main outcome of the workshop is that this goal has been realized. In particular it was stimulating to see junior participants making new contacts across the two communities. Chitta Baral, Thomas Bolander, Hans van Ditmarsch, and Sheila McIlraith will be the organizers of a to be proposed follow-up Dagstuhl workshop. An epistemic planning webpage will be launched by Thomas Bolander.

Organization/Format: Much encouraged by repeated interaction with the Lorentz

Center staff, the organizers scheduled a large amount of time for discussion groups. Apart from the invited talks, two rump sessions of each 90 minutes (18 presentations in total) were also scheduled on Monday and Tuesday, for which participants could sign up on the first day of the workshop. The originally planned five discussion groups were on Monday made into three, after first asking the participants to sign up for interest. The `planning' discussion group (formerly `strategy') produced a taxonomy of epistemic planning methods. This investigation may be continued into a survey for the journal special issue. The `concurrency' discussion group delivered a detailed comparison table, identifying (technically) the commonalities and differences in the approaches, thus offering specific directions for filling gaps and providing formal ways of comparison. The discussion group `agency' was changed mid-week into a discussion group `gossip', given participants' interests. Such flexibility in reshaping the program during the week was very effective to capture the participants' attention. The participants wanted time on Friday afternoon for one-on-one discussions with collaborators. This was enabled by an impromptu decision to shift the Friday afternoon session to the end of Friday morning.

Social programme The boat trip was the high point of the week. The guided tour of

of the Boerhaave Museum and the subsequent Public Lecture also went down well. A very welcome and helpful gesture of the Lorentz Center staff was to encourage us to organize an evening of epistemic games (for which epistemic planning is important) in the Lorentz Center on Tuesday afternoon (`encouragement' means: free beer, drinks, and nibbles).