Environmental resources are the backbone of government economy. Most of the existing systems are failed to provide a secure management and maintenance of passing emergency information to the worker during risk conditions with the coverage of wi-fi technology. Data communication through radio wave is not possible on far end of subversive mines because asymmetrical data proliferation and inadequate rate of recurrence scope 3 kHz-300 GHz. Most injuries taking place in subversive mines include rock falls, omissions, and bangs. The mining confinements remain pretentious by lung sickness due to inhaling dust and toxic gas, which are emitted from mining environment. Hence, we are formulating this issue with novel Li-Fi technology. It has covered a wide range of frequency (430-790) THZ. In this paper, we propose a LDM (Light-based Decision Mechanism) based HSM (Human Safety Management) algorithm to evaluate and monitor the abnormal conditions through an IoT sensor data. A precise machine learning based decision-making computational system is mandatory to evaluate the comfort level conditions and drove the notifications to the management remotely to overseer the mining environment conditions. This system estimates emitted toxic gas and evaluates the level of effect during the time of mining processes. This method requires unidirectional communication; therefore the LiFi channel is absolute suitable to accomplish secure and noise less data distribution within a nano second ample of time. The experimental result shows the performance of our algorithm for better than prevailing data analysis algorithms. © 2019 IEEE.