A Multi-Disciplinary Approach of Business Architecture and its Business Intelligence Applications for IoT Big Data

摘要

The purpose of this paper is to clarify a multi-disciplinary approach for deep learning based IoT (Internet of Things) application architecture in business. In IoT, collaboration among domain knowledge experts, data scientists, systems architects, and total system consultants are inevitable. Because each specialty is totally different and it is hard to cover all of knowledge in one person. The role of consultants is important how they can coordinate these multi-disciplinary people and knowledge, and manage to create architecture for domain specific IoT applications, such as healthcare, autonomous driving, smart city, and smart home, etc. We show how these multi-disciplinary people interact, design and implement business oriented deep learning applications. In many cases, even domain knowledge people cannot tell where such data are hidden in the companies and how should be analyzed. In business applications, to show rapid prototyping system as Proof of Concept (PoC) is also important for these multi-disciplinary people and business executives to discuss the effectiveness and direction of data analysis. Especially business executives sometimes only can know how IoT analytics systems are effective only by showing the analytics results in intuitive graphical expression. We show some example applications applied to this multi-disciplinary method and evaluate its effectiveness.