Air Taxi, Regulation

Interview: How Uber Views Electric VTOL Aircraft, UAM Operations and Safety

By Brian Garrett-Glaser | January 20, 2020
Send Feedback

Jacek Kawecki, vehicle components lead at Uber Elevate, discussed how Uber views eVTOL aircraft, UAM operations and safety.

Uber’s future aerial rideshare product, which it expects to launch commercially in 2023 in three cities, has gained significant ground in the past few months. The addition of automaker Hyundai and Joby Aviation — now flush with $590 million in investment and a partnership with Toyota to assist on manufacturing — as vehicle partners brings their total to eight, adding significant credibility and monetary backing to the Elevate project.

Uber has taken an asset-light, steward-of-the-ecosystem approach to developing aerial ridesharing vehicles, setting targets and requirements for participation, based on their understanding of what aircraft will safely and profitably perform the urban air mobility mission. Many of these fundamental beliefs were defined in the Elevate white paper released in 2016.

With hundreds of companies making progress on eVTOL aircraft, unmanned traffic management systems and other elements of the future aerial ridesharing ecosystem, Avionics International interviewed Jacek Kawecki, vehicle components lead at Uber Elevate, to further understand Uber’s assumptions and approach to eVTOL aircraft development, operations and safety.

Can you share the most recent version of the “complexity-criticality vs performance” graph Uber has presented when discussing eVTOL aircraft? Where do the Nexus 4EX and Hyundai S-A1 place?

The most recent version of the complexity-criticality vs performance (CCvP) graph was presented in Jacek Kawecki’s 2019 summit presentation on eVTOL Safety, Criticality, and Complexity. As a policy, we do not comment on our manufacturing partners’ vehicle performance.

A slide presented at the Elevate Summit in 2019, charting aircraft complexity and criticality against performance. (Uber)

You said at Uber’s Elevate Summit in 2019, “We believe that the best aircraft for Uber Elevate operations will have minimal criticality but a balance of complexity and performance.” Could you further explain this?

We believe intelligently applied redundancy can provide the best mitigation against maintenance-induced accidents. We have encouraged our manufacturers to consider vehicle designs that minimize the use of critical parts, and by extension, incorporate the fewest potential single points of failure. Examples of critical parts in aviation today would include rotorcraft cyclic, collective, and tail rotor mechanisms.

As demonstrated by Uber’s vehicle requirements and the CCvP graph, you clearly don’t consider multicopters an option for Uber rideshare — yet companies such as EHang, Airbus and Volocopter are fervently pursuing their development.

Is there a world or a mission set where multicopters make sense?

Uber’s network development strategy has been developed through a comprehensive analysis of urban mobility and rider preferences (such as price/time tradeoffs).  This network strategy has informed our vehicle performance requirements on metrics such as speed, payload, range, and acoustic compatibility. Multicopters are not presently competitive with vectored thrust or lift + cruise vehicles on our network.  Outside the Uber network, there may be short range markets where a multicopter is the right choice.

Uber has outlined an approach to safety that targets a 10-8 chance of catastrophic failure for aircraft — a step down from what is required of commercial airliners, as well as EASA’s special condition for eVTOL aircraft released last year — while focusing on improving operational safety through a layered approach that aims to reduce pilot error.

Will this be sufficient to satisfy the public and regulators, given the historically high bar presented by commercial air travel?

Uber looks to US Part 121 commercial transportation as our north star for actual levels of safety.  To achieve this goal, we are looking holistically at the contributing factors to fatal accident rates for existing Part 135 operations and the areas we can make the most significant improvements.

Target Level of Safety (TLS) is a far more nuanced concept than is often publicly presented.  TLS targets are applied to aircraft functions through the system safety assessment process, with top level targets applied to each of the potential catastrophic failure conditions on the aircraft.  Multiplying an expected number of flight hours by TLS and estimating failure rates is an overly simplistic and inaccurate method of predicting expected accident rates.TLS also fails to capture the fact that over 70% of fatal accidents are attributed to operator error (inclusive of pilots, maintainers, dispatchers, ground crew).  This is true in every operational context from Part 91 to Part 121.

You also commented at Elevate 2019, discussing vehicle airworthiness requirements: “We don’t believe that today’s airworthiness requirements should be based on future projections of growth.”

Can you elaborate on this?

We predict extraordinary growth in UAM operations over the next decade.  However, even our most optimistic projections don’t show UAM flight hours per year exceeding private or commercial transport rates until well after 2030.  Considering these projections, and the lower inherent risk of small airplanes with limited occupant seats, we don’t see value in mandating transport category Target Level of Safety (TLS) targets.  That being said, we fully expect vehicles operating on our network to have far higher mandated safety targets than conventional 4 passenger + crew vehicles which are already operating over urban environments.

Another slide presented by Uber at the 2019 Elevate Summit. (Uber)

How does Uber plan to keep the occurrence of maintenance-related safety issues to a minimum?

As mentioned in a previous answer, we believe one key piece to reducing maintenance-induced failures is to reduce the critical parts count and incorporate intelligent, fail-functional redundancy.  We also believe that this approach permits reliability centered maintenance (RCM) methodologies. By moving away from critical parts with fixed time-between-overhauls (TBO), you can move toward on-condition maintenance. Ideally, a component is kept in service until its performance as long as its performance is within acceptable safety levels.

This has a direct connection to safety, according to Dr. Saleh at Georgia Tech, between 14 and 21% of US civil rotorcraft accidents from 2005 to 2015 were attributable to flawed maintenance and inspection. Among these accidents, 31% occurred in the first 10 flight hours following maintenance, suggesting a strong trend of maintenance induced infant mortality.  Reducing contact maintenance may help to reduce maintenance induced failures. Additionally, we believe electric aircraft will utilize HUMS (Health Utilization Monitoring Systems) to help maintainers identify failing or damaged components. HUMS data can add an extra layer of protection against mechanical systems failures.

Uber Air aircraft will be held to the same standards for establishing maintenance processes as existing aircraft 135 operations. We believe that in addition to using a safety management system (SMS), existing regulatory oversight of MROs will work for Uber Air.

Receive the latest avionics news right to your inbox