In this research we explore utilizing IoT for safe and reliable decision-making. We aim to devise and analyze an optimization model to maximize the lifetime of Internet-of-Things (IoT) energy-constrained networks in the presence of attacker compromise and environmental failure factors. We consider emerging IoT technologies, including the Low Power Wide-Area Network (LPWAN) and mesh networks, where an IoT cloud manages the deployed IoT network adapting to sensed phenomenon changes in the deployment area retrieved from peered cloud-based environmental monitoring systems. Such IoT technologies can be utilized in security sensitive emergency response and alert systems. While increasing redundancy of IoT to cloud communication increases reliability and fault tolerance, it also comes at the expense of increased energy consumption. We consider an optimization problem using IoT energy models and communication patterns to determine the best redundancy level to be applied in the IoT network such that the lifetime is maximized while achieving the communication reliability requirements in the presence of attackers and unreliable sensing environments. We demonstrate the feasibility of our approach. Depending on our findings, we could pursue other avenues of IoT energy-constrained fault-tolerant and security related research topics which could span various topics such as smart city services, Trust Management, and Social IoT aspects (satisfying research priorities).