Grouping similar buckets

My aggregations buckets are sorted by doc_count at the moment. I would like to group them by similar buckets (e.g 40 km/h, 50 km/h, 60 km/h) even if doc_count is different. Is this even possible?