Dana Fisman

Runtime consultants (RCs) provide online recommendations toward satisfying temporal objectives, and in this sense resemble path-planning methods. While RCs emphasize formal guarantees and many path-planning approaches rely on heuristics, both settings use quantitative objectives to guide sequential choices. This work bridges these largely separate communities, showing that both face objectives that fail to yield decisive recommendations and address them via quantitative refinements, which improve solution quality and induce notions of progress. We illustrate this perspective for reachability, generalized reachability, and limit-average objectives.