Managing Complexity: A Resilience Enabler
As I’ve said multiple times, a significant portion of investment in resilient space systems is applied to specific threat mitigation approaches, perhaps at the expense of other important elements. Development of software and hardware to counter specific threats is vital, however without other less obvious investments the introduction and operation of these mitigations can become problematic. This is especially true as the complexity of space systems continues to grow. Both governments and commercial operators have responded by taking advantage of the cost reductions in computing power to add automation to their mission management systems to augment human operators. This has largely been applied to day-to-day operations for often very static system configurations in relatively benign environments. When things begin to change quickly the first response is often to more directly engage the system operators when the situation extends beyond canned automation programming. But for the most part the automation is sufficient to ensure baseline system performance and mission plan execution.
When considering resilience implications several of these assumptions fail by definition. The introduction of threats implies a more rapidly chaining environment and the need to diagnose system performance compromise by these threats. System operation has thus veered from the daily operational stance and into what may be unfamiliar territory. To be sure, some responses may be forecast and provided as options to the automated assistant. But given the gravity of the potential loss of space assets if the wrong response is executed, humans are usually quickly invited into the process. Often the operators are confronted with a vast amount of data from which they must divine the preferred mitigation for the threat, including the recovery of system capabilities.
The intersection of system capabilities, threats, and available mitigations (including concept of operations) can multiply to produce a large number of potential outcomes. This complexity gives rise to a fundamental challenge in achieving a highly resilient system: quickly analyzing available information, including situational awareness, identifying and characterizing the threat(s), and then determining the best course of action by which the system users experience the lowest impact. Mitigation may involve multiple techniques, executed on a specific timeline, in an optimal sequence in order to achieve the desired system response. It can involve any or all of the four key resilience attributes: avoidance, robustness, recovery, and reconstitution. And as the number of capabilities, threats, and mitigations grow, so does the solution space, which must be evaluated quickly to assure the highest resulting resilience.
The obvious path forward is to evolve advanced automation to move it into the realm of artificial intelligence, leveraging machine learning, to help the operator rapidly sift through the myriad options at their disposal to meet an operationally relevant timeline. Furthermore, this AI assistant must be given both authority and access to enable it to, under proscribed circumstances, orchestrate mitigation effects as quickly as possible and engage and inform the human operators as it does this. This orchestration may include command, control, and configuration of satellites, ground sites, and user terminals, all within a short period of time. It may also require the AI to interface with other systems, such as indications and warnings (I & W) to obtain necessary situational awareness data.
However it is achieved, management of system complexity, including operations, is becoming more and more important to resilient space system design and should not be deemphasized. Rather, as mitigations are developed, they should be compatible with a greater resilience orchestration function, whether it is AI-driven or not. Only through such an approach can the full potential and value of threat mitigation features be fully unlocked to provide maximum system-level resilience.
