Fig. 1 illustrates the research design adopted in this study. First, a systematic literature review of concepts and theoretical frameworks on energy management practices and the IoT paradigm was conducted. This included critical evaluation of IoT definitions, technologies and factory applications. For this purpose, several keywords were used in the search process, such as “energy management practices,” “energy efficiency in production,” “energy monitoring,” “energy consumption awareness,” “energy data in real-time,” “smart meters,” “IoT technology,” and “IoT and energy efficiency.” Related papers were found using search engines, including Google Scholar, Web of Knowledge, Elsevier, and Scopus.
Second, given the nature of this research paper, a qualitative research approach based on semi-structured interviews was used: the interview format provided a level of structure in order to cover some main topics, but left a certain degree of flexibility by allowing for follow-up questions in order to provide clarification (Saunders et al., 2009). Namely, six executives of Technology/Solutions Providers were interviewed: two general managers, two sales managers, one account manager and one product manager. These types of companies are regularly in contact with their customers and provide services for them, such as storing customers' data in the cloud and analyzing such data: the interviewees can thus be classified as experts in energy management practices. Interviewing experts is commonly used in the literature, as in Koskela (2011). The interviewees were asked different questions to highlight the state of the smart meters, sensors, and applications they offer to customers. In addition, they were asked to define their customers' practices in energy management after installing smart meters and collecting and analyzing energy data. These questions were guided by the literature review performed in Step One, and provided insight into the current sustainable practices at manufacturing companies and opportunities for improvements based on the availability of energy consumption data. Moreover, following the methodology in (Koskela, 2011), four further interviews were conducted with industry professionals. Two of them were experts in IoT technology; they were asked questions focusing on IoT technology and its applications for energy management. The other two were energy management consultants, who were asked about current energy management practices and integrated energy data in production decisions at the production level. Third, we collected information available online from ten manufacturing companies that have already adopted IoT technology for energy efficiency. The information collected included which technology had been installed and which energy management practices were adopted after collecting and analyzing energy data, and which benefits they observed.
Eventually, relying on the literature review and interviews, inductive modeling was adopted to build a framework for IoT-based energy management in production, so as to define how energy information could be integrated into production management decisions. In order to test the validity of the framework, it was reviewed by three energy management consultants (two experts from the interviewees mentioned before, plus an additional third reviewer).