Context-aware systems are crucial in modern cyber-physicalsystems (CPS), enabling dynamic adaptation to changingconditions. Many such systems involve complex variability.Modelica, a leading equation-based language for modelingand simulating physical systems, struggles to manage thiscomplexity. As variability management becomes more complex,traditional Modelica constructs such as conditionalstatements, state machines, and state graphs becomeincreasingly inadequate and difficult to maintain. Thiswork introduces a context-oriented modeling paradigm forModelica by integrating Petri Net-based context controlwith FMI-based Variable Structure Systems (VSS).Specifically, we present Context Petri Nets (CoPN) as aformal mechanism for representing and managing contextualdependencies. By combining CoPN with VSS, our approachenables advanced variability management, supporting themodeling and simulation of complex, context-aware CPS.