Workflows in Microservice Based System of Systems
Abstract: The welfare of society has increased together with the developments of new technologies to ease manual work and provide affordable goods for everyone. Now, in the era of computers, society demands products with the same quality and variety as those formerly produced by craftsman, but with the price, manufacturing time, and quantity of mass production.As a response to the demand for mass customization of products, industry is leveraging the adoption of new technologies to improve their manufacturing processes. Research on computer technologies has kept a fast pace, new prototypes and tools can be created swiftly, and changes to software systems can be applied at anytime and anywhere provided the system is connected, usually to the internet. Almost every person in the planet interacts daily with a device fitted with an embedded computer, whether is a smartphone, a car, or a smart building. A sensor and a microprocessor can add value in many situations, nevertheless the key for societal change seems to be the digital communications, in what has been coined Internet of Things (IoT). Industrial products need to endure under harsh operation conditions, therefore additional requirements are set for Industrial Internet of Things (IIoT). At industrial plants, the goal of adding new devices is to precisely control remotely and digitally the physical manufacturing process. The virtual management of the devices through digital communications, and their behaviors interdependence with the real process, has been capture by researchers in the concept of Cyber-physical System (CPS).By themselves, IIoT and CPS, could already spur a revolution where thousands of new devices will populate smart factories. However, it will probably hinder manufacturing as current engineers lack the ability to manage the countless number of devices, handle the quantity of data provided, or use the information in advantageous business decisions. Thankfully, progress in big data, cloud computing, machine learning and artificial intelligence can help with these new issues, altogether starting the fourth industrial revolution that researchers have predicted.In this thesis, we explore solutions to the problem of management and integration of heterogeneous software systems present in Industry 4.0 use cases. The research is based upon systems designed according to a microservice, or even a Service Oriented Architecture (SOA), and grouped in dynamic System of Systems (SoS).The presented solution consists of a workflow management and a workflow executor system working in tandem to select, implement and execute workflows at run-time. This new system is motivated by the lack of capabilities in basic microservice development frameworks and the capability to distribute the autonomous execution operation from centralized systems in the ISA-95 standard manufacturing automation architecture.
This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.