Commercial and military systems have evolved into complex Systems of Systems (SoS) that incorporate advanced and diverse technologies. Associated with these SoS are emerging behaviors that require decision support well beyond the capacity of human reasoning alone. To fill this gap, Artificial Intelligence and Machine Learning (AI/ML) can assist SoS Engineering with respect to operationally realizing the full potential (e.g., speed, scale, and accuracy) of the capabilities offered by these SoS. This presentation provides a review of the concepts of complex SoS and emergent behavior, discusses the fundamentals of AI/ML, and then ties these together to demonstrate the role of AI/ML in SoSE through specific commercial and military use cases. These use cases include:

Object Recognition and Detection Enhancement via Reinforcement Learning Yield (ORDERLY). ORDERLY is an AI/ML-based capability that supports SoSE by autonomously screening massive collections of sensor data from SoS and transforming this raw data into actionable information. Disaggregated Distributed AI Chat Enabler (D2ACE) System. D2ACE applies AI/ML techniques to Chat-based SoS. D2CRaB supports SoSE by correcting spelling, typos, and other corruption in chat messages, recognizing uncommon language formats, and autonomously prioritizing and reducing the quantity of chat messages to only those relevant to specific objectives and intent for a given mission. Distributed Disaggregated Communications via Reinforcement Learning and Backpressure (D2CRaB): D2CRaB is an AL/ML capability that supports SoSE for Communications. D2CRaB introduces two new advances to address problem of effectively communicating within distributed and disaggregated operational environments. These new techniques resolve the immediate congestion issue and then assist with maintaining congestion free network traffic. In summary, through this background information and these use cases, the audience will emerge from this presentation with a focused understanding of AI's role in SoSE with respect to supporting human decision-making for complex, emergent SoS.

Advances in technology - distributed systems, digital engineering, big data, cloud computing and of course, artificial intelligence - are all profoundly affecting today's systems both the systems themselves and the approach to engineering these systems. Systems of systems are similarly affected - and the impacts in some cases may be amplified for systems of systems given their complexity, emergence, and evolving functionality/behavior. New technologies may lead to new or unexpected system architectures, patterns or, in the case of AI, evolving behaviors. This panel will look at the impacts on systems of systems from the introduction of new or advanced technologies. This includes the challenges the technologies bring to SoS and considerations for managing the challenges.

Telepresence, the extension of human ‘presence' beyond their immediate physical surroundings, has significant implications for a multitude of possible futures. In almost all scenarios, we will witness the emergence of a telepresence economy and, more broadly, a telepresence society. When this presence extends to virtual worlds, it is commonly referred to as ‘telexistence,' encompassing extensions of our presence through both physical and virtual avatars. With remote presence, truly ‘the sky is the limit,' as activities expand beyond Earth's surface. A space economy has indeed been ignited, albeit currently limited to Earth's orbit (envisage communication and Earth-observing satellites, and sub-orbital flights as the seeds of space tourism). A lunar economy, or more specifically, an economy in cis-lunar space, is expected to follow within the next 2-3 decades, driven by space robots and telepresence. These space robots will possess high levels of autonomy and, eventually, superior intelligence. However, until that advanced stage is reached, they will be teleoperated or tele-supervised by humans. Beyond being instrumental in completing tasks, human interaction is crucial, as machines will learn from human demonstrations and exemplifications, a process I refer to as ‘interpolative intelligence.' This talk will project us into some of the possible futures in which telexistence will profoundly transform our society. A new fabric of systems of systems is thus woven, as the singular space of operation for an individual extends to a multiplicity of remote spaces, concurrently, or more realistically, through time-multiplexing, as our abilities for concurrent processing are, at least for now, rather limited.

AI has jumped to the forefront of our collective psyche in the past year. In some ways this looks much like the early stages of the typical technology hype cycle. But we also see indications of something quite unique, something that will transform the way we work unlike anything else in recent history. As Yogi Berra and Neils Bohr both observed, making predictions is difficult, especially about the future. Instead we will look at impacts already happening, unique characteristics of AI that make it unlike other technology waves, identify some of the well-known and lesser-known risks, and attempt to extrapolate a bit about the next steps in Systems Engineering with AI and for AI. We will end this thought-provoking session with a discussion of the most critical obstacle to the application of AI in Systems Engineering.

The panel will debate these topics in the settings of complex, emergent SoS for both commercial and military operational environments. To guide the discussion, the panel will:

• Provide examples of technologies in which SoSE can apply AI/ML to support human decision-making including:

• Autonomous Vehicles
• Communications
• Command and Control
• Object Recognition and Detection

• Discuss pros and cons of using AI/ML for these applications
• Identify gaps in proposed AI/ML solutions
• Identify tools and methods with which SOSE can help meet these challenges
• Suggest areas for future work