<p>Industrial wireless networks (IWNs) have gained considerable attention for their ability to deliver time-critical services, which are enhanced by device-to-device (D2D) communication that reduces transmission delay and offers advantages like scalability, flexibility, and low cost. In IWNs, age of information (AoI) is the significant metric, which is used to measure the real-time performance of data delivery from the perspective of the destination node. In this work, the Fractional Puzzle Optimization Algorithm (FPOA) is proposed to train the global critic with priority-aware scheduling. This model consists of local actors for D2D users and an edge-assisted critic network. Here, edge-assisted critic aims with its global coalition structure generates output for each actor after scheduling. Here, the proposed FPOA is formed by combining the Puzzle Optimization Algorithm (POA) and Fractional Calculus (FC). The FPOA-based priority-aware scheduling of IWN is analysed with various metrics like fairness, fitness, data rate, power consumption, transmission time, and throughput, which obtained superior values of 0.888, 0.809, 0.909, 0.209W, 0.558&#xa0;s, and 47.157&#xa0;Mbps. The devised model cannot be applied to large-scale industrial applications as it has increased power consumption and also the performance is affected due to the number of resources and devices. The devised FPOA-based priority-aware scheduling has improved the reliability and efficiency of the model and the use of edge computing reduces latency by providing real-time responses. The high performance of FPOA makes it well-suited for priority-aware scheduling in IWSN.</p>