Insights for Prototyping for the Industrial Internet - Iterative Prototyping Loops of Connected Sensor Platforms
MetadataShow full item record
This thesis aims to find out insights for how new concepts for Industrial Internet can be developed in the early stage product development projects. The Industrial Internet is complex because it is mixing multiple disciplines. A product development team needs to master software, electronics, mechanics and data science in a connected environment. The research objective of this thesis is to test out multidisciplinary iterative low-cost prototyping in early stage Industrial Internet projects in order to collect and derive both practical and theoretical insights for industry and research to use. It does it both positivistic way, i.e. what did work based on real experiences from the six projects in this thesis, and also what did not work and what problems were found based on the same projects. The purpose of these insights is to provide a part of toolbox of technologies, tools and methodologies for developers in the industry and engineering design to tackle this complexity and make better products that achieve their purpose. The steps to achieve this objective is to hypothesize about how to prototype Industrial Internet projects, to showcase how to create new concepts for the context, and present their potential value through proofof- concept prototypes as information for investment decisions considering to build complete systems. The thesis seeks to find insights for the three research questions: • RQ1: What are the technologies, tools and methodologies that enable Industrial Internet prototyping in the projects of this thesis? • RQ2: How can Industrial Internet be prototyped based on the projects of this thesis? • RQ3: What are the practical and managerial implications of prototyping the Industrial Internet in the projects of this thesis? Results used in this research are based on real data acquired from real company and research problem contexts of early stage development projects that the author has done or taken part in. The project data, which we define as technical development, process development and the project outcomes were recorded using various methods, such as creating case studies, creating hypotheses based on case studies, interviews, notes, prototypes, project logs, photographs, videos, documentation and reports. The captured data was analyzed and categorized on each project and, if deemed applicable, published in peer-reviewed conferences and journals. The insights and recommendations can be seen as the results of this thesis and they include: 1. Combine low-cost prototyping tools, e.g. laser cutting, 3D-printing and cardboard models, with sensors to produce data to learn about the problem and test the possible solutions, which could be used to make decisions and explore Industrial Internet applications. 2. In the early stage product development context, collect data that might seem redundant. Practice has shown that if the developer is not completely sure what he or she wants to do with the data and how, it is better to collect everything in raw format without pre-processing or filtering the values. 3. Use mechanically low-resolution prototypes in conjunction with electronics to create test datasets. Mentally hold the opportunity still open for changing the whole physical setup while iteratively using the explorative data analysis as a feedback to the physical implementation as well as the software design if prototyping leads to new insights. 4. Make a pre-project (an extended prototyping phase) before the investment decision - especially with heavy industry, in expensive and complex cases. The conclusions of this thesis include how the two main enablers for prototyping the Industrial Internet are firstly the new user-friendly technologies and tools, such as Arduino, that were developed late 2000s, helping to prototype the choice of core technology for projects and secondly, much more sophisticated toolbox of methodologies on process level to address unknown problems (rapid prototyping, iteration cycles, agile, scrum, design thinking, etc.). During the projects while prototyping low-resolution physical concepts combined with relatively high resolution sensors technologies and exploratory data analysis turned out to be a powerful combination for developing early stage Industrial Internet devices: This combination enables faster learning and yields innovative product concepts. The gist is to apply Wayfaring method to connected products and to iterate with the data and create new ideas based on insights from the data analysis. The main point for prototyping the Industrial Internet products is the intent of putting the decision maker in a better decision position. Rather than fixing the requirements and the architecture directly, we are proposing an extended prototyping phase in order to come up with better decisions that are based on understanding more of the problem and the solution space, and eventually understand what is a good solution in its context. It is in key importance to understand which variables are actually having an effect in the project. The purpose of prototyping is to reduce uncertainties in the project, e.g. which parameters we are dealing with. As the result of the extended prototyping phase, we should have a proofof- concept prototype and a lot more information about technologies, costs and opportunities to make the real investment decision for the R&D project.