The aviation community has an unprecedented opportunity to take safety to a whole new level. A digital transformation of society—a fourth industrial revolution—is creating some profound challenges and opportunities for aviation safety.
Billions of devices around the globe are continually producing immense volumes of data. And rather than functioning in a stove piped manner, they are part of a collaborative ecosystem, exchanging data in real time with each other and the broader environment.
In the aviation domain, new commercial aircraft generate nearly one terabyte of data per flight—twice that of their predecessors. Meanwhile, the number of airborne vehicles is growing. And as the number of connected devices grows, the number of potential interactions grows exponentially.
The vast potential of this ubiquitous data lies in combining it with advanced technologies such as artificial intelligence/machine learning (AI/ML), cloud computing, the Internet of Things (IoT), and advanced wireless technologies to drive innovation.
Evolving safety in the digital age
Today's aviation system is the safest in history. However, traditional safety approaches may be difficult to utilize in an interconnected world. The accelerating changes in the scope, scale, speed, and complexity of the aviation environment will require new approaches to aviation safety.
MITRE, operator of the US Federal Aviation Administration's (FAA) federal R&D center, proposes a way forward with its “Next Level of Safety” approach.
“With our world becoming more complex and dynamic, safety needs to evolve to proactively identify hazards before they manifest into accidents,” says Greg Tennille, managing director of transportation safety at MITRE. “Traditional safety practices, such as spot checking and ad-hoc coordination, must be augmented by new approaches to ensure public safety in this rapidly changing environment. We believe adoption of our Next Level of Safety approach can shift the paradigm of aviation safety across the globe.”
Four principles define the Next Level of Safety: a system-of-systems approach, proactive monitoring and analysis, the sharing of safety information across stakeholders, and smart policies focused on safety outcomes.
Gaining a holistic view
The Next Level of Safety requires that the entire system – people, process and technology be accounted for when considering safety, including the increasing role of human-machine teams. Adopting a system-of-systems perspective encourages a shared, comprehensive understanding of the environment.
To achieve this common understanding, stakeholders can use safety management system principles to engage in proactive risk management. Under the Next Level of Safety, safety management systems (SMS) shift from the traditional single-organization focus to a 360° safety review that incorporates all stakeholders. This allows multiple organizations to analyze and manage their common risks, share discoveries before accidents or incidents occur, and understand the effectiveness of mitigations. Systems level safety management also enables the community to discover better solutions and offer faster feedback than any one organization operating independently.
Consider the interactions between airlines, the airport and air traffic control, where each is operating with an SMS. All three are analyzing data to spot new hazards, such as pilots frequently entering the wrong taxiway at a confusing intersection. Each is in self-discovery and self-correction mode, finding and fixing these issues. Working separately, one organization might find and make a correction that masks the issue for a time. That correction may address the root cause within one organization but not for the system of systems. If multiple organizations work together to monitor leading indicators, the opportunity to spot new issues before they create significant risk is increased.
In the future, much of the cross-organizational sharing will move beyond raw data to include AI-generated findings regarding “what to look for”—in essence, patterns for leading indicators of safety issues. The organizations can then work together to improve the depth of the analysis and determine the best solution.
Likewise, airlines and maintenance organizations are key sources of information for manufacturers about what is or isn't working on their aircraft. Perhaps one maintenance shop finds a better way to correct a problem than is provided in the maintenance manual. By sharing that best practice with other maintenance organizations, the manufacturer, and airlines, all can benefit.
In its US aviation oversight role, the FAA could facilitate the adoption of such practices. For instance, airlines could be required to communicate identified issues and fixes to all relevant stakeholders, and the FAA's audits of their SMS could serve as an enforcement mechanism. A similar approach could be taken in other parts of the world.
Systems level approaches also enable decision making about safety to be integrated with decision making about efficient operations. The way in which airspace and air routes are designed and used must reflect a balance of safety objectives and economic factors such as fuel consumption, passenger value of time, and fleet utilization.
In addition, as airspace and airport capacity are dynamically adjusted to account for scheduled space launch and re-entry operations – or to accommodate a growing volume of drone activity – techniques such as dynamic flow management and collision risk mitigation outcomes can be simultaneously achieved.
Delivering in-time mitigations
Today's increasingly data-rich environment is also enabling the shift from a retrospective to a predictive safety mindset. Proactive monitoring and analysis that employs advanced technologies can detect anomalous events and emerging hazards, provide insights and predictions, and allow for appropriate actions to be taken before they manifest as accidents or incidents. Under the Next Level of Safety, stakeholders combine a sophisticated network of fused operational, systems, and environmental data with AI/ML-based technologies to generate timely safety intelligence in both strategic and tactical timeframes. Human decision makers and autonomous systems working together as a team use this information to automatically monitor system performance, understand operational impacts, and identify unanticipated system interactions. Long-range forecasts of accident risk inform strategic safety investment priorities. Near-term forecasts provide tactical support to operational actions.
An example of this principle in action is Aviation Risk Identification and Assessment (ARIA), an automated capability the FAA deployed in 2020 to analyze surveillance data from the entire US airspace system in near-real time, identify events that might have posed a safety issue, and prioritize them for deeper investigation.
ARIA takes a risk-based approach to safety analysis. It uses defined metrics to flag any airborne or surface event that represented a safety concern—even if the operation was in compliance with rules and regulations. This provides FAA quality assurance officers with a more complete safety picture than is possible with compliance-based review processes alone. ARIA also looks for patterns in the events so that systemic risks can be identified and addressed.
Expanding sources and perspectives
A key element of the Next Level of Safety is data democratization, the process of ensuring that data and information are accessible to as many stakeholders as possible. Shared data enables stakeholders to produce more robust predictive analyses and safety intelligence insights, coordinate their mitigations, and understand how to recover from unforeseen hazards.
To maximize the benefits of a systems perspective, members of the community must be incentivized to share safety data, from local to global levels. As a foundational step, issues related to data security, governance, trust, and standards need to be addressed. For example, a set of common frameworks and data standards would facilitate the broad adoption of interoperable solutions and accelerate the impact of data sharing.
Within the FAA, efforts are already underway to develop a global safety information management platform that could support multiple international partners in sharing data for safety analysis.
Meanwhile, government regulators can serve as a catalyst in the fostering of information sharing among industry stakeholders. This includes sharing insights from regulatory monitoring activities as feedback to assure the effectiveness of safety controls. Trusted third parties can assist in building trust between government and industry stakeholders.
In the USA the Aviation Safety Information Analysis and Sharing (ASIAS) program, a partnership between industry, the FAA, and MITRE, has helped to dramatically improve safety in the US airline industry. As the aviation community pursues the next level of safety, this model could be applied more broadly, such as regionally or globally.
A focus on safety outcomes
Because traditional safety regulatory approaches are often insufficient in today's increasingly dynamic and complex digital environment, government regulatory policy is also shifting. It's becoming more performance-based and more focused on safety outcomes. It's also promoting industry self-discovery, disclosure, and correction. This approach enables innovation by granting industry flexibility on how to achieve the desired safety outcomes.
From a Next Level of Safety perspective, regulators take an integrated approach, supporting the oversight responsibilities of all organizations in a “shared mission” operational chain i.e., manufacturers, maintenance organizations, operators, and regulators.
Adoption of a “safety continuum” is an example of this approach. Not all aircraft and airborne operations, for example, need to be held to the same costly requirements to maintain consistent safety outcomes. Depending on the aircraft's mission and operational needs, different safety level requirements can be established. A safety continuum aligns regulator and industry safety management goals with the public's safety level expectations, while also granting industry innovation flexibility.
On an operational level, with the smart policy approach, regulators use oversight to promote greater shared understanding of the system level operational environment, potential risks and associated impacts among stakeholders, in addition to determining regulatory compliance. Oversight attention is focused on the shared discovery of risks and the better coordinating of the layers of mitigation across stakeholder organizations.
Moving forward as a community
The fourth industrial revolution is transforming the aviation operating environment by creating more connectedness than ever before between systems and human users. Understanding the safety implications of this greater connectivity requires broad community understanding and engagement, with a more diverse set of users, to generate new safety insights.
To achieve this vision, members of the safety community must come together and accelerate implementation of the Next Level of Safety principles on a global scale. By taking a systems perspective, sharing data, proactively monitoring and identifying risks, and implementing smart policy, we can create a safer world.