Talk:Multi-agent pathfinding

Review

Hi, I read your article and corrected some typos. I think it is really clear and well written, and I found the topic really interesting. I have just a couple of suggestions:

1) you could rephrase the last sentence of the introduction, which for me is not totally clear ("However, given the applications in which MAPF is involved such as automated warehouses and airport management, it is important to reach a trade-off between the efficiency of the solution and its effectiveness.")

2) in the paragraphs "Prioritized Planning" and "Optimal MAPF Solvers" you talk about MAPD, but only later on you explain what they are: is it a typo or is it done on purpose? In this case, you could consider introducing them before, because otherwise it might be a little confusing.

Good job! Fraapal (talk) 13:27, 1 July 2022 (UTC)

Thank you for the suggestions! I will rephrase the sentence and correct the MAPD typo! BenedettaFlam (talk) 14:36, 1 July 2022 (UTC)

Review

Nice article! I've just checked some typos but, other than that, good job! Andre lleo (talk) 14:46, 1 July 2022 (UTC)

Observations and suggestions for improvements

The following observations and suggestions for improvements were collected, following expert review of the article within the Science, Technology, Society and Wikipedia course at the Politecnico di Milano, in June 2022.

The article is interesting and clearly written. The main point concerns the terminology about graphs: as correctly stated in the backgound matherial, a graph has nodes and edges. Later in the article, edges become arcs. Arcs are used in the context of oriented graphs, and this is confusing since, at the beginning, the graph is assumed not oriented. Moreover, when discussing the types of collisions, (a) and (b) have to be swapped, in my opinion: the vertex conflict generally happens at time x but not at time x-1, so that the agents positions are different at time x-1 (see figure (b)); on the other hand, the edge conflict assumes that the vertex conflict happens at time x-1 and at time x for the two agents (see figure(a)). Finally, when defining the action function, the node v' has to be adjiacent to v, and different from v (not from v'). The only typo I found is "udually" at the end of Types of Conflicts section.

-- BarettoDiArchitettura (talk) 10:39, 12 September 2022 (UTC)