<p>Improving the accuracy and efficiency of order delivery, the logistics sector now has the technical basis for route optimization with real-time scheduling systems, thanks to advancements in intelligent technology. Big data analysis is the technological instrument that this article suggests using to better enhance intelligent logistics’ working circumstances. In addition, the essay concludes with experimental results comparing the strategy presented here with other popular approaches. Intelligent logistics route optimization with real-time dispatch system development is both made possible by big data analysis technologies, as is seen from the clear experimental gap.Using DRL, we develop a pedestrian route planning system in this study. Our route planning is based on pedestrian traffic predictions on the road network, and we utilize journey time consumption as our measure. we provide an implementation of a short-term traffic stream forecasting system that makes use of LSTM. One deep learning method, LSTM, can learn both linear and non-linear traffic patterns, as well as long-term relationships. The results showed that the proposed LSTM–DRL model had a better prediction capability reduction in MAE, MAPE, and RMSE to 8.32, 19.9%, and 5.65, respectively, whereas the LSTM errors were 10.33, and the ARIMA errors were 11.79, and show that LSTM–DRL model performed even better with distances traveled. This also demonstrates how effective the model was at optimizing errors related to travel-time prediction estimates. Ultimately, the hybrid model conclusion offers a sophisticated and exact method of logistics route optimization in moving dynamic scenarios.</p>