We present adaptive refinement approaches for meshfree methods, drawing parallels and contrasts with traditional mesh-based adaptivity. Meshfree collocation methods stand out for their ability to refine computational domains without the constraints of mesh connectivity, offering a level of flexibility and efficiency. Unlike mesh-based methods, meshfree refinement inherently avoids issues such as hanging nodes, simplifying the refinement process. With an emphasis on automated adaptation for unsteady fluid flow simulations and applications in the automotive industry, we showcase the use of a-posteriori error indication techniques. This talk highlights how insights from the extensive work on mesh-based adaptive refinement can be effectively adapted and tailored to meshfree methods.