Machine Learning for Predictive Analytics in Autonomous Vehicle Supply Chain Management

Abstract

[1] [2]The traditional supply chain management (SCM) model becomes very costly with substantial inventory and long delivery time in the context of supply chain’s overall operation, thereby facing various business challenges in the evolution of the digital economy. It has been already demonstrated clearly in some recent research works that introducing intelligent vehicle technologies has shown improvement in the overall performance of the supply chain by integrating new technologies into conventional transportation management methods, involving cooperative transportation, and integrating increasingly advanced technology into the specific transportation fields. However, intelligent vehicle technology provided by traditional supply chain management is inoperative to drive faster, safer, dependable, and greener strategies, for example, virtual reliability, trust, adaptation, mobility, privacy, and others in both networked and cooperative technological futures. Just like any other emerging disciplines, autonomous vehicle supply chain management (AVSCM) in a blockchain-rich environment urgently needs to be driven by the research ideology of “digital, electric, networked, shared” development stage, rather than simple replicating road-tested supply chain management.[3]Except for the surroundings of vehicle networks, machine learning has been broadly used with the purpose of maximizing the operations of diverse industries and systems. For instance, some accomplished researches incorporate techniques of ML to enhance retailing and production (Miles and Pavone 2012), medicine (Alpaydin 2010), service quality (Park et al. 2010), travel and transportation (Jiang et al. 2011) and other fields. Accompanied by the continuous progress of skills and methods, human knowledge can be studied better and even replaced by the help of connatural intelligence (Riera et al. 2016), which has brought unprecedented and comprehensive development to different acts and schemes. After collecting a number of achievable results in numerous following tests and attempts, a fact we have to understand is that the unsung heroes of all these remarkable successes are a kind of accurate and good interactive mechanism that is called “machine knowledge” or “machine learning,” which diverts computers and networks seem more clever as well as adroit and confidential by executing vast numbers of possible carbonate computions.