Our Story: Part 3

In this post we continue our timeline-themed recap of the internal monthly updates circulated throughout Airbus as Vahana was built and tested. This year in particular was huge for the Vahana team as it marked the first flight and subsequent major flight milestones accomplished by the hardworking, dedicated group.

We know you’re excited to continue reading the timeline BUT if you haven’t yet explored our post from 2017 we highly recommend you start there. Check it out here then come right back!

For those who are up to speed - enjoy 2018!

January + February 2018

  • On January 30th, the FAA Safety Evaluation was successfully completed, resulting in a Special Airworthiness Certificate for the first Vahana aircraft N301VX.
  • On January 31st, thanks to a break in the weather, Vahana completed its first flight consisting of a 53 second hover at five meters. The following day, a similar test was conducted to validate some specific safety features.
  • Throughout the rest of the month of February, the team worked to upgrade the motors before moving on to additional flight testing.

March 2018

  • This month was focused on completing characterization and integration of the MAGicALL motors which offered high efficiency, about 50kg of overall weight savings, and cleaner integration.
  • The final sets of landing gear were completed and passed drop testing with the fairings integrated.
  • Alpha Two now has landing gear, motor mounts, and completed wiring.
  • Finally, the whole team was thrilled to demonstrate complete transition from hover to fully wingborne flight on the subscale.

April 2018

  • In April, Alpha One’s new motors were extensively tested (including low voltage, high voltage, and power-on regression tests) in preparation for the FAA Safety Evaluation in May. This evaluation will approve the aircraft for flight and expand the existing operational envelope.
  • The first version of sense-and-avoid software, including the deep convolutional neural network (DNN), was deployed to Alpha Flight Computer. This will allow validation of the DNN in the real vibration environment of the aircraft.
  • It was this month that Fast Company named Airbus among the finalists for World’s Most Innovative Companies, specifically mentioning Vahana in the honor.

May 2018

  • At the end of April, the team completed all integration and testing tasks needed after the propulsion system upgrade and on May 1st, the FAA SAfety evaluation was completed. This opened the door for the team to complete the following tests:
    • FT03 OGE Hover: Completed on May 2nd, this regression test was used to verify basic functionality of all systems in hover out of ground effect. FAA representatives were present to witness this test.
    • FT04 Hover Endurance: Completed on May 3rd, this OGE hover test lasted approximately 4.5 minutes and gave the team insight into the performance of the motors, controllers, and batteries. Overall, all systems performed wonderfully.
    • FT06 Chirp Test: Completed on May 8th, this test applied chirp inputs to some of the outer loop inputs. The vehicle responded as expected and validated our nonlinear models.

September 2018

  • Throughout the summer, the team completed nearly 30 test flights and dozens of test points. These tests were focused on performance verification (critical azimuth, ascent/descent rates, gust envelope expansion) and system characterization (failure management, chirps).
  • Additionally, our new propulsion system meant our final FAA Safety Evaluation, which expands our operating area to its maximum extent, and regression testing to verify the performance of the VPFs. Once this is complete, transition testing will begin starting at 20 kts forward speed and increment in units of 10 kts until the stall speed is exceeded.
  • In parallel with flight testing, propeller truck testing has been performed over a matrix of propeller speeds and crosswind angles to validate the VPF performance models. Additionally, the sense-and-avoid landing zone evaluation system has been demonstrated on a surrogate vehicle.

October 2018

  • Vibration testing and additional flight testing brought some unforeseen issues that made the month of October particularly challenging. That said, as with all testing, much has been learned and solutions are in place to resolve the outstanding issues. And, after all, it’s better to find issues on the ground than in the air.
  • The greatest challenge has been with the lift fans (notably the most complex component on the entire aircraft). These issues resulted in a delay of a few weeks.
  • Other tasks that have been worked in parallel to resolving the above issues include continued development of the UAM concept of operations, technical evaluation of self-piloted operations risks, and configuration exploration.

November 2018

  • Subscale Omega completed its 60-knot equivalent flight and quickly followed with fully autonomous transition.
  • On the full-scale aircraft, all motors were overhauled, tested, and rebalanced on our test stand before being integrated on the aircraft. After completing ground tests and regression flight tests, due to the incredible work of the integration and test team, we were able to begin the transition test campaign beginning with the 20-knot transition chirp test.
  • In parallel with these efforts, we have been focused on integrating ourselves into the newly introduced Airbus Urban Mobility organization.

December 2018

  • Once the 20-knot tests were complete, the results were compared to our performance models to identify any unusual behavior. As the team prepared for the 30-knot flight test, those models were used to verify that our fully autonomous algorithms (flight controls, navigation, failure detection and mitigation) were performing as expected in both our hardware-in-the-loop simulator and in our Monte Carlo analysis, which is used for identifying off-nominal performance and overall test point risk (e.g. off-runway landing).
  • Simultaneously, the subscale was used to test the next flight test point including evaluation of motor failure performance in flight.
  • With the required inputs complete for the 30-knot test, and despite a long run of poor weather (typically heavy fog), on 19 December we flew the 30-knot flight test.