Archive | 2021

“Where does that module live?” The difficulty of extending SOSA into smaller, expendable applications

 
 

Abstract


The Sensor Open Systems Architecture™ (SOSA) Consortium seeks to decrease costs and increase performance of large-scale, standalone sensor systems like multi-sensor Electro-Optical Infrared (EO/IR) or Radio Detection and Ranging (RADAR), typically hosted on manned aircraft platforms. However, the greatest benefit of the SOSA Reference Architecture—interoperability—could also fuel the future of small, expendable systems such as Unmanned Aerial Vehicles (UAVs) and wearables. Not surprisingly, a sensor architecture crafted for large systems can be a challenge to host on small or distributed hardware. This paper presents three examples of the appeal and challenges inherent in defining an open architecture for the world of small sensors by asking, “Where does that module live?” The first example presents the case of sensors on small UAVs; for this case, the SOSA Technical Standard defines classes of connectors, in the vein of SAE AS6169™A, as well as mechanical interfaces. Interoperability and standardized interfaces would make sensors for UAVs easier and cheaper to assemble and deploy. However, the current SOSA Components and Modules definitions can be awkward for small UAV sensors. For instance, a small sensor may not contain a processor for hosting Automatic Target Recognition (ATR) on the sensor, but instead the platform may contain that processor, hence distributing the SOSA Modules onto different hardware elements on and off the host platform. The second example presents the case of using VITA 74.0-complaint hardware (VNX) for wearable sensor applications. It is appealing to apply the SOSA Reference Architecture to this new paradigm, but this tiny form factor as a part of the Internet of Battlefield Things (IoBT) may require a new contextualization of the boundaries of a reference architecture that includes distributed composable hardware elements. The third and last example presents the case of a swarm of sensor-enabled UAVs. The swarming sensor concept encourages yet a further reconceptualization of the SOSA Reference Architecture boundaries and the SOSA Data Model. As an open architecture originally intended for standalone systems and sensors, the SOSA Reference Architecture naturally assumes a sensor communicates primarily to the platform and never directly to other sensors. However, in order for a swarm of UAV-based sensors to reach maximum effectivity, the SOSA Reference Architecture could potentially extend its scope to include a swarm class with SOSA Modules that might not live either on the platform nor in a single sensor—they could be shared entities or organized in a hierarchy. These three examples illustrate how a forward-looking sensor open architecture can evolve to drive the future of smaller applications by expanding into these novel realms to provide a smart, flexible, small-form factor architecture.

Volume 11753
Pages 117530B - 117530B-19
DOI 10.1117/12.2585950
Language English
Journal None

Full Text