Abstract
Arrays, or datacubes, form a key data abstraction next to sets, hierarchies, and graphs. They appear in manifold ways in practically all areas of science, engineering, and business. While datacubes are established well in OLAP, they are less common in other domains. Practically, such arrays always constitute Big Data, and so database support for arrays becomes increasingly important. The rasdaman system has pioneered array databases, with its array query language adopted into SQL as Part 15: Multi-Dimensional Arrays. With application domains in science and engineering (as opposed to business analytics) the system is in operational use worldwide. We present the current state of the system, with its key innovations in model and architecture, illustrate its impact on standards, and give an outlook on future directions.