SBIR-STTR Award

The objective of this proposal is to research and develop computer-vision based smart harness capable of retrieving packages of up to 300 lbs. Our eVTOL platform can then transport these items up to 10 miles.   While our SBIR 19.3 Phase I focused on overall feasibility, and deemed it feasible, this application zeroes in on an important functionality that has as yet to be researched and developed: the smart harness.  Namely, what recognizes the package and what retrieves it?   In the past, humans have largely been used to connect harnesses and assure safety of cargo. In addition, helicopters (expensive, loud, dangerous) have been used to then transport the cargo. Our hover-primary eVTOL platform is incredibly maneuverable. Optimized for safety with built-in redundancy, it can be trusted to carry precious cargo.    Technically, the challenge might at first seem straightforward. But the challenge remains, as ever, in the integration and development of custom elements that need to work well in a variety of situations. We aim to implement a computer vision software algorithm to recognize a package, design a custom attachment for said package, and develop an electromagnetic attachment system that will hold the package.   Think of your old Macintosh MagSafe power adapter or the charger for the Apple Watch. Click, and you’re ready to go. The goal, ultimately, is to build in further autonomy to the platform so that it can intelligently move large objects without human supervision. To begin, we must be able to locate, identify, and capture the package properly and reliably.   This STTR, in cooperation with the University of Texas at Austin, aims to create a design and build a first-of-its-kind prototype that can perform these activities.

In our Phase I effort, we worked with the University of Texas at Austin to research an important computer vision question, 'How can an autonomous system correctly identify a unique marker from above?' While itself a non-trivial question, the answer begins to the open the door to an important, larger set of technical questions. How can we then integrate that computer vision system into a larger airframe? How can we have that airframe fly in a stable and precise enough manner to successfully crane in and out cargo? How can that system integrate both crane and camera systems into an algorithm that could, one day, allow for fully autonomous cargo drop and retrieval? Answering those questions is the goal of our Phase II. The Phase I had scope enough for the computer vision question. We seek to use the Phase II award to further flesh out this vital research and development effort into a prototype with critical software and hardware integrated into our triply-redundant flight computer. It is our goal to take the fruits of this technical effort and create an operational aircraft that the DoD could use to fulfill the 'Logistics Under Attack' Concept of Employment. We will discuss this more in the anticipated application and benefits section below. The technical goal here is to research, develop, and employ an autonomous computer vision and crane system into our unique electric vertical takeoff and landing aircraft platform, HEXA. We are honored to continue working with the University of Texas at Austin (whose aerospace engineering program currently ranks 7th in the United States) on this important initiative.