Large safety-critical real-time systems are typically created through the integration of multiple components that are developed mostly independently from each other. This creates a challenge for the timing verification of both the independent components and the integrated system as a whole given that traditional real-time scheduling techniques required full knowledge of the taskset and resources of the whole system contrasting with need of component independence. The challenges stemming from this situation can be categorized in three broad areas. First, at the system decomposition stage it is necessary to fix the component timing interfaces for each of the components in order to ensure the schedulability of all the components and to enable each component-development team to define their own tasksets and perform internal schedulability test. A key tradeoff in this area is efficiency (schedulable utilization) vs internal flexibility. Flexibility, in particular, is a key aspect at this stage when only incomplete information on component tasks, period, deadlines and most importantly execution time (among other parameters) is available. Secondly, safety-critical systems demand some form of certification. In this case, compositional technology should be able to provide human-processable assurance arguments for component-integration timing correctness and component isolation while at the same time providing sound component schedulability techniques. Finally, during the system evolution new compositional technologies should enable us to isolate the effect of changes not only around the current decomposition structure but also allow us to reshape the decomposition structure or provide multiple levels of composition to limit the scope of the impact of the changes. CRTS invites papers that describe not previously published state-of-the-art research, work-in-progress, or suggest open problems covering one or more of the topics of interest to the workshop. Submissions should not exceed 8 pages in two-column, single-space, 10pt format (ACM SIG proceedings templates / SIGBED Review submission guidelines) or two pages in extended abstract format for work-in-progress / open problems. A submission implies permission for ACM to publish the paper in print and digital formats. Authors retain copyright. Accepted papers will appear in a special issue of ACM SIGBED Review.