Wireless sensor networks (WSNs) have become increasingly important in the informative development of communication technology. The growth of Internet of Things (IoT) has increased the use of WSNs in association with large scale industrial applications. The integration of WSNs with IoT is the pillar for the creation of an inescapable smart environment. A huge volume of data is being generated every day by the deployment of WSNs in smart infrastructure. The collaboration is applicable to environmental surveillance, health surveillance, transportation surveillance and many more other fields. A huge quantity of data which is obtained in various formats from varied applications is called big data. The Energy efficient big data collection requires new techniques to gather sensor-based data which is widely and densely distributed in WSNs and spread over wider geographical areas. In view of the limited range of communication and low powered sensor nodes, data gathering in WSN is a tedious task. The energy hole is another considerable issue that requires attention for efficient handling in WSN. The concept of mobile sink has been widely accepted and exploited, since it is able to effectively alleviate the energy hole problem. Scheduling a mobile sink with energy efficiency is still a challenge in WSNs time constraint implementation due to the slow speed of the mobile sink. The paper addresses the above issues and the proposal contains four-phase data collection model; the first phase is the identification of network subgroups, which are formed due to a restricted range of communication in sensor nodes in a wide network, second is clustering which is addressed on each identified subgroup for reducing energy consumption, third is efficient route planning and fourth is based on data collection. The two time-sensitive route planning schemes are presented to build a set of trajectories which satisfy the deadline constraint and minimize the overall delay. We have evaluated the performance of our schemes through simulation and compared them with the generic enhanced expectation-maximization (EEM) mobility based scenario of data collection. Simulation results reveal that our proposed schemes give much better results as compared to the generic EEM mobility approach in terms of selected performance metrics such as energy consumption, delay, network lifetime and packet delivery ratio.