To implement a resilient manufacturing system, adaptive systems are required to respond to changes and disruptions. In this article, we present a research prototype of an adaptive production system. Adaptivity is implemented here in three components and on different levels. The first component is an adaptive robotic system that can be easily trained to pick up heterogeneous parts and place them on an autonomously guided vehicle. This component implements adaptivity at the task level. The second component is a planning system that can be used to reschedule production orders on an ad hoc basis. This system also allows the simulation of machine failures to analyze resilience. This job shop planning system implements adaptivity at runtime at the level of production execution. The third component is a modular process modeling and execution system that enables adaptivity at the process level. It supports the users in redesigning production processes with the help of a graphical user interface. The overall system is modular, with the three components being run in an adaptive, agile, and decentralized way. It forms a system-of-systems that shows resilience to several disruptive events.