The NASA X4+ Advanced Air Mobility program: The Inclusion of Weather Information in a Unique Live Flight Test

As a follow on to the 2021 X4 simulation series, part of the NASA National Campaign (NC1) for Advanced Air Mobility (AAM), the North Texas Cohort, a collaboration of academic institutions and private companies operating in the Dallas/Fort Worth metroplex, performed a year long experiment to include and demonstrate additional complexity within the larger NASA/FAA concept of operations. A significant advancement from X4 was the inclusion of live updating and test case weather information within multiple subsystems. Data sources included high resolution gridded model data from the experimental HRRR and the RTMA, assimilated multi-model point forecasts, rapidly updating Doppler weather radar data from Nexrad and from the CASA DFW network, and warning products from the NWS WFOs. The culmination of X4+ was a live flight test using a Bell 407 helicopter as a proxy for an EVTOL aircraft, with simulated weather constraints closing a flight track during the planning stages, and closing a vertiport in-flight requiring an on-the-fly reroute. The flight test, planned long in advance, happened to occur in the presence of ongoing showers and weak thunderstorms and live data extractions indeed did occur and were distributed throughout the system during the experiment in the prescribed manner. Weather representations have been included in the system in several ways– Firstly, in accordance with UAS Traffic Management (UTM) standards, as UAS Volume Restrictions (UVRs). Here, specific weather tolerances have been assigned to flight events, with active monitoring and extraction of contoured polygonal geofences from the relevant observed and forecast data soureces. Upon extraction, the encapsulating 4d volume of the UVRs were entered into the system's Discovery and Synchronization Service (DSS), compared with user-registered subscription volumes, with the details subsequently delivered to individual Providers of Services for UAM (PSUs) when applicable overlap was found. Meanwhile these constraints were maintained by the publishing 'privileged' PSU and subject to be queried on demand in accordance with UTM policies. Herein we provide suggestions for metadata relating to weather contained within required DSS subscriptions to ensure relevant weather info is being extracted and delivered to users. Moreover we benchmark performance for a DSS system with may weather UVRs, and suggest the spatial scales of both weather representations and the footprint of airspaces as well, so as to be able to rapidly retrieve, perform overlap calculations, and deliver georeferenced descriptors to users. Secondly, we introduced a flight winding service, a python based REST application using gridded winds aloft from the HRRR. This service is used to better estimate waypoint arrival times in advance of takeoff from Operational Intent data, a sequence of 4d trajectory points describing a planned flight operations. The winding service uses the initially planned waypoint locations, predicted arrival times, airspeed velocities, and related track speed limits, and then advects that path through the gridded modeled windfield with an implementation of the Bresenham 3d algorithm. It calculates the segment relative net head/tail winds and crosswinds. The service is capable of adjusting planned air speeds, within track and vehicle limits, to meet given waypoint times, or to use given airspeeds and recalculate waypoint arrival times. In some cases air speeds cannot be safely adjusted to meet times, resulting in both being speeds and times adjusted. Moreover both the headwind and crosswind components are used to adjust safety buffers for tactical deconfliction purposes. Thirdly, we have introduced a vertiport weather service making use of multi-model point forecast data. In addition to standard meteorological products (Temperature, Dewpoint, Wind Speed, Wind Gusts, Ceilings, Visibility) it also provides precipitation likelihood percentages, and winds at different heights above the vertiport. This service is used in several ways. For our Demand Capacity Balancing (DCB) product, we dynamically adjust the number of vehicles that can take off and land in a given timeframe based on the characteristics of the weather. DCB inclusion is a unique implementation compared with traditional blocking UVRs. Moreover, to generate Operational Intents, our flight planning tools adjust the departure and approach sequences around vertiports such that they are primarily into the wind to match standard commercial air traffic procedures. Finally, within the program have tied in our weather related services into an advance health and monitoring system, with simple liveliness and quality assurance checks. In simulations we have intentionally triggered service outages and registered errors. The downstream airspace management services periodically monitor the health and status API to selectively enable or disable contingency management routines, including changing how departure and approach paths are chosen, increasing buffer size for conformance monitoring, and dynamically adjusting capacities at vertiport locations and crossing points. We believe that to date these tests are unique among sanctioned AAM demonstrations using current concepts of operations, and herein we further discuss lessons learned, techniques, and recommend ways that weather information might continue to be utilized and distributed within system modules (DSS, PSUs, Air Track descriptors, etc). We see this work as entirely complementary to the ASTM weather standards for sensing, modeling, and quality assurance that are being developed in parallel.