Vertical flight is not for the faint-hearted

Today’s urgent need for green transport and noise-free aircraft is poised to propel growth of the electric vertical and take-off (eVTOL) market.

Over the last decade, innovators and legacy manufacturers have been quietly developing the core technologies and testing prototypes and demonstrators. These include startups such as Joby, Archer, Beta Technologies, Lilium, and Vertical Aerospace, as well as established aerospace leaders including Airbus, Boeing, and Rolls-Royce. In 2021 alone, these companies raised around $7 billion in private investment; more than doubling the amount over the last decade.

Vertical flight is very capital-intensive, with significant R&D costs in the emerging technologies on which it relies: high-density batteries, distributed propulsion systems, and novel aerodynamic designs which improve aircraft performance without compromising system redundancy or safety.

Nonetheless, some organizations have succeeded in overcoming major engineering hurdles. That is an important step forward. The next few years will now be about proving the safety and operational reliability of their aircraft to airworthiness authorities and to the public.

How do we prove and certify novel aircraft designs?

A major endeavor when bringing a new aircraft to market is certification. It include ground testing, simulations, in-flight data acquisition, and critical software testing, and detailed data collection and reporting (see Appendix 1 below for more details). This process demonstrates how the aircraft systems meet EASA/FAA airworthiness requirements.

Certification for a new, low-carbon aerospace system is by far the most expensive and challenging task prior to market entry.

Such processes follow a rigid and gated approach known as Validation and Verification (V&V). In this process, requirements are first validated and cascaded down to the component level of the aircraft. Then, during the design, build, and test phases, the product is verified with a set of analysis and flight tests agreed with regulators. These aircraft design reviews are known as Preliminary Design Review (PDR) and Critical Design Review (CDR), after which the certificate will be released.

Here a glimpse of how this looks:

Today, the frontrunners in the nascent advanced air mobility space are racing to complete their aircraft design phase, which includes a concrete certification plan. In addition, most developers are also seeking to bend the cost and time curves when compared to traditional aircraft development, while following robust certification programs.

Digitalizing Next-Gen Aircraft Certification Workflow to speed time to market

As with so many things, digital technologies can help speed and optimise complex processes, whilst also increasing rigor

At Capgemini Engineering, we have identified four significant areas of digital acceleration, where we see next-generation aviation companies achieving high standards in their design-for-certification goals with quality, time, and risk reduction:

1. Digitalizing the certification workflow to improve data traceability, from regulatory airworthiness to evidence compliance, whilst enabling automated VTOL test case generation and systems trade-off analysis
2. Incorporating data-driven machine learning to automate test scenario generation
3. Automating task orchestration, and the generation of test cases combining a wide variety of sources of evidence, such as Model-in-the-Loop (MIL), Software-in-the-Loop (SIL) and Hardware-in-the-Loop (HIL) testing, computational analysis, and flight tests
4. Building digital twins of flight scenarios. This will also support high-volume manufacturing and fleet operations that rely on flight test campaign results.

A digital certification process will help ease the burden of aircraft certification, while automating test reviews and compliance evidence generation.

Here is a glimpse of how a digital certification process could look:

The technology of test automation and digital certification

Doing all this means choosing, deploying, configuring and learning new technologies. And organizations face more technology decisions than ever. From smart cloud modernization, to AI and machine learning, to data analytics, weaving together different technology platforms can seem overwhelming.

There is no one-size-fits-all approach. But we present here a few examples of technologies that are helping companies in delivering safe and reliable aviation products with reduced lead times.

Constrained Software Test (Test Case Generation):

Based on established principles of statistical testing, it enables companies to:

• Establish early verification conditions (VCs) based on requirements and systems specifications
• Define which software test to perform
• Perform random tests derived from VCs and system constraints
• Perform checks of the Systems Under Test (SUT) behaviors against reference models and VCs

Validation Plan Generation and Intelligent Testing (ATLAS Test Scheduling):

AI-based generation of a “Complete/Explicable/Smart” Validation Plan using genetic algorithms & Swarm intelligence to optimize data coverage:

• Generation of hundreds of test scenarios in a matter of seconds
• Automatic generation for «homogeneous covering» of test cases
• Automatic validation of plans and updates after each change based on test results
• Improved faulty or limited case diagnosis

Auto-detection of Flight Test Anomalies (Improved Result Assessment and Classification):

Detection of non-standard flight characteristics can be automated using a combination of statistical methods, Principal Component Analysis (PCA) and machine learning. Our proprietary AI-enabled anomaly detection has been used to save thousands of hours each year, whilst enabling deeper data analytics and diagnosis.

Validation & Verification of Flight Controls (Automated Simulation and Analysis):

Automated V&V processes that accelerate the Validation & Verification of flight controls and handling qualities. Thousands of simulation results that were previously written and checked manually are now automatically generated and analyzed on the complete set of certification requirements. Our customers are now able to define their own templates and automatically fill their test results including the release of certification dossiers.

How Capgemini can help: Faster digital certification, without compromising rigor

When it comes to speed, design, and certification, digital process that enable test automation, process traceability and better data capitalization are paramount. Companies looking to shape the future of aviation will need to collaborate with multiple suppliers and partners and leverage digital platforms.

Capgemini Engineering is a long-standing engineering and R&D partner to aviation, working with suppliers and government regulators for many decades. We understand what it takes to design and certify parts and systems for the safety-critical aviation sector and the automotive and railway industries.

The deployment of model-based systems engineering, digital twin, and the digital engineering practices we have put in place to support certification activities for our customers have been demonstrated to improve teams’ productivity and reduce recurrent costs across a product’s lifecycle by up to 40%. And we proactively invest and develop in the core technologies that will ensure a better and more compelling future for present and new generations.

Appendix 1: The path to certification

Currently, as part of the aircraft certification process, companies are required to:

1. Engage as early as possible with regulators such as FAA or EASA to collaboratively define how airworthiness standards will apply to manufacturers’ specific eVTOL architectures and systems design. This includes:
   • Definition and agreement of working methods used for the development and certification of the aircraft
   • Agreement of the certification programs and level of involvement from the regulators
2. Define a test plan that includes ground testing, simulations, in-flight data acquisition, HIL and SIL (Hardware- and Software-in-the-Loop) and critical software testing. For example, addressing:
   • How should manufacturers acquire the information required to comply with Part 23|Special conditions (SC) VTOL requirements?
   • How to optimize test plans based on data availability and flight test campaigns
3. Collect data and engineering artifacts from real and virtual mission testing, including:
   • Production and collection of existing engineering artifacts and models, including model fidelity analysis
   • The collection of flight handling and performances data from HIL/SIL simulations, flight tests and engineering analysis
   • Automate the search and annotation of data to match the informational needs
4. Perform compliance and systems performance checks:
   • Display the collected information in an accessible and automated form 
   • Assess and classify tests results   
5. Evaluate compliance and submit documentation to airworthiness authorities