Features of communication barriers in oil and gas logistics

The key industry – specific features of oil and gas logistics are the high complexity of supply chains and the multi‑level management system, which determine the quality and characteristics of the communication process between industry participants. An important aspect of the interaction between participants in supply chains is the presence of communication barriers that hinder the effective exchange of information. The study of information logistics flows in the oil and gas sector, the main communication barriers have been identified: organizational, technological, informational, and human. Their substantive characteristics and the degree of their impact on the effectiveness of logistics processes have been clarified, which made it possible to systematize measures to overcome them into two groups: technological and organizational. This also made it possible to justify the use of artificial intelligence technologies as a tool that enables solving complex tasks of managing logistics flows based on high‑speed information processing and multifactor analysis. Identifying the specific features of transactions in the oil and gas industry – the uniqueness of the range of services and the geography of supplies, the high cost, and the significant level of uncertainty – allowed us to conclude that it is advisable to use a predictive support system for negotiations and communications in the context of negotiation processes in oil and gas logistics. This system combines machine learning methods, natural language processing, and predictive analytics into a single platform integrated into the company’s existing digital ecosystem. The mandatory conditions for designing this system are defined as follows: diagnosing the current state of communication processes and identifying problem areas aimed at establishing critically important functional and non-functional requirements for the development of a predictive support system for negotiations and communications. It is proposed to use a multi‑level predictive analytics model that includes three methodological layers: historical analysis – processing of archived data on completed transactions; behavioral analysis – forming indicators based on communication patterns in the current transaction; contextual analysis – taking into account external factors; a substantive description of the methodological layers is provided.