I. INTRODUCTION
Advancing the real-world potential of autonomous robotic systems by leveraging heterogeneity [1] is an increasingly pursued area of research. As standalone system classes, Mobile Manipulation Systems (MMS) have been extensively deployed in exploration, construction, and inspection [2]–[5], while Micro Aerial Vehicles (MAVs) have been utilized for search and rescue, surveillance, industrial inspection, and broader commercial activities [6]–[11]. Exploration of unknown and unstructured environments remains a core mission objective with numerous algorithms developed [12]–[16]. However their focus remains on finding exploratory paths which follow collision-free routes achievable by a single agent –the exploring one– therefore disregarding edge cases where an agent may only fit through an opening safely with the help of a second robot. To this end, we argue that a heterogeneous system-of-systems robotic platform can gain entry into otherwise inaccessible areas, by autonomously detecting such marginal ingress opportunities (gaps or apertures) in an unstructured / partially-collapsed environment, and deploy a different class of miniature robot through them by leveraging high-precision mobile manipulation.