Automated Flight Safety Inference Engine (AFSIE) System

We propose to develop an innovative Autonomous Flight Safety Inference Engine (AFSIE) system to autonomously and reliably terminate the flight of an errant launch vehicle. This proposed phase 1 research is innovative in that it combines proven NASA-developed AFS algorithms, real-time hazard assessment algorithms and hazard envelopes generated from Joint Advanced Range Safety System Real Time (JARSS RT) and an on-board vehicle simulator into a refined onboard software inference engine that monitors navigation states, mission flight rules and onboard anomaly instrumentation. An autonomous flight safety system must be able to reliably perform accurate and autonomous navigation so as to determine the vehicle position, velocity and attitude states in real time. Reliability requirements for AFS are high due to stringent loss-of-life constraints, often leading to redundant navigation sensors with attendant cost impacts. Our innovative solution proposes to satisfy RCC accuracy and reliability requirements by exploiting the low-cost COTS sensor and processor architectures that are currently being baselined for the Common NanoSat/Launcher Avionics Technology (CNAT) study and a Nano launch vehicle avionics design. This dual use hardware implementation will greatly reduce the recurring costs for the production of an autonomous flight safety system. This has significant implications for reducing the costs for launch vehicles, particularly Nano and Micro Satellite Launch Vehicles (NMSLV), where range safety costs currently consume a burdensome percentage of the launch cost. Under this proposed phase 1 effort, we will 1) identify the range requirements and develop a plan for range safety for approval of the system, 2) identify reliable low-cost COTS hardware that satisfies the range accuracy and reliability requirements and, 3) develop an end to end simulation to demonstrate the AFSIE Concept of Operations.