Strengthening the Long-Term Robustness of Artificial Intelligence: A Nature-Inspired Solution Sustaining Value Network Resilience

Developing a well-performing artificial intelligence model is hard - employing it productively is even harder: Among many other aspects, application circumstances and associated data inevitably change. Without proper maintenance strategies, the model’s robustness is at stake. Given the prevailing advancement of artificial intelligence in productive use and the absence of elaborate strategies, this serious problem has become increasingly omnipresent. We outline possible solutions and demonstrate these within real-world business use cases from industrial research.