Lean based cycle time reduction in manufacturing companies using black widow based deep belief neural network

Abstract

Manufacturing companies in India are looking forward to innovative principles in order to come out with novel product designs to compete with other companies. Recently the lean principle has been focused on many companies for maximizing profit through minimizing waste. So, to track and eliminate unwanted activity, various researchers have developed several approaches based on the lean principle. When the manufacturing business tries to improve efficiency, cost base, and customer responsiveness, cycle time is one of the viable parameters that should be optimized as much as possible. This systematic study focuses on reducing cycle time in the manufacturing industry to boost productivity. However, effective tracing and elimination of waste with an improved, streamlined process are not achieved using the traditional lean approach. To solve this problem, a lean approach based on the deep learning technique is designed in this current research to minimize cycle time. Initially, the current state mapping of an organizational structure is done using Value Stream Mapping (VSM). Then, Modified Deep Belief Neural Network is used to detect excess time consumption. The modification is done by optimizing the weight parameter with the assistance of the black widow optimization algorithm. Once the excess time is detected, it is further eliminated to maximize the profit. The simulation analysis of the proposed BWDBN approach is done, and the actual value is compared with the predicted value. The analysis indicates that accuracy, precision and recall of 98%, 97% and 96.5% are reached using the proposed BWDBN. This suggested that novel product design and maximized profit can be attained using the proposed BWDBN approach.