An efficient safest route prediction-based route discovery mechanism for drivers using improved golden tortoise beetle optimizer

Several commercialised route recommendation systems only consider the metrics like cost, time, and distance. The essential metric 'safety' is neglected by the existent systems. It suggests only the short way and doesn't include any safety information, such as crime awareness, road availability. This paper describes an inventive ideology to discover the safest route with minimal risk score for security of the road travellers. Hence, a new safety route navigation mechanism is developed to solve the challenges in the traditional route discovery approaches using deep learning. In the developed route discovery mechanism, the examination of the safest roads is done by the developed deep learning network, where the network is trained with the inputs obtained from the roads, such as road surface conditions, Road users, weather conditions, traffic conditions, accidental cases, and crime areas. The availability of the safest route will be determined by the 'Long Short-Term Memory with Attention Mechanism' (LSTM-AM). The route discovery is done with the help of a developed 'Fitness-based Golden Tortoise Beetle Optimizer' (FGTBO) with multi-objective constraints like distance, time, and road availability. The implementation outcome of the developed route discovery scheme will be validated with the traditional route discovery approaches concerning various measures.