Digital Certification of Fast, Lightweight Autonomous Platforms

Fast Lightweight Autonomy (FLA) is very challenging to evaluate and certify by humans especially when the drone falls into a new category that is yet to be defined and classified by the DoD and it operates on high speeds that is challenging for humans to safely deal with risking the possibility of crashes.  At Istari, we believe that accelerating the certification of FLA is a key area required to accelerate innovation and enable a new category of defense technologies, ensuring compliance and keeping an edge in the field. Simulating and ensuring that autonomy meets certification is a must given the increasing complexity of autonomous systems and the vast variety of possible environments that an autonomous system might operate in which makes it really difficult to truly be able to test autonomy effectively in the physical world. End-to-end simulation has the potential to completely change how new systems are designed, tested, and fielded.  Simulations are becoming a standard part of creating products and through our subsidiary Top Flight Technologies we have created a preliminary version for a larger drone we would like to see if we will be able to do the same for smaller drones and for FLA. By being able to do this digitally we will be able to change the paradigm of how certification happen as instead of building physically then testing which is costly and time consuming, we will be able to test multiple times in the digital realm before needing to build thus cutting down time and cost significantly thanks to simulations that might be able to do the same work that typically happens through the physical certification process. Additionally, physically testing FLA is impractive due to the difficulty in measuring FLA by a human being and the vast scenarios in which a drone should be tested for autonomy as cost to risk ratio of doing physical tests is at least 2x higher than doing the tests in a simulated environment. Other players in the space are either trying to build a walled garden “Dassault as an example” or are a non-profit that doesn’t move fast enough “OSLC as an example”, in our case, we want to build a healthy interconnected and fast growing ecosystem that is missioned to increase the fidelity of accuracy of simulations to ultimately replace physical tests.