Self-adaptive IoT

IoT is just beginning a decade-long development – The IoT service market will reach maturity in 2025 and keep growing strongly until 2030. The critical challenge in IoT systems is to achieve service levels and secure the uninterrupted flow of business-critical data; emphasis is on the speed of making new business data available as soon as possible.The critical challenge in IoT solutions is that they are subject to inherent uncertainties in their operational contexts. Moreover, IoT resources are also constrained with compute power, sensing, communication, battery levels. So operational contexts and IoT resource constraints cause a significant challenge to make this solution available 24*7. A key question is can we address all these operational context challenges during development time? Not really; these uncertainties are difficult to predict during IoT blueprinting, which would sometimes result in Over/ Under provisioning of the resources. The resource demands of IoT applications fluctuate during run-time purely due to their event-driven nature.This paper details an approach to handle those challenges. It starts with how IoT service management differs from regular IT Management. It then details what challenges IoT deployment faces and its impact on the sustainability of the solution. Finally, a self-adaptive approach is explored, i.e., the ability of a system to adjust its behavior in response to the perception of the environment and the system itself. The option is based on the usage of the self-adaptive system using the machine learning approach.