The research is focused on exploring and solving the issues that affect a Digital Twin (DTw) operating as a mimic of an active on-the-ground physical process. We present new machine learning (ML) technologies and modern strategies to meet the capability gaps identified in four DTw literature reviews. We set up simulations and tested DTw's ability to supervise local and multiple remote sites. Target domains include healthcare, fire safety, security, traffic, and a safecity.

The paper presents insights gained over two years, working on an Australian Commonwealth Government-sponsored research project based on a Smart City ontology. The research focused IoT ecosystem enables the study of the range and scale of services that can be delivered to a multifaceted IoT environment. Custom-engineered, Sensor nodes, generate parametric measures from fixed or mobile locations, 24hours a day, for 365days a year. Case-1, with public health and safety as the key objective, as an example and demonstrates how lessons learned were applied to other use cases, using a flexible and configurable IoT ecosystem. A Semantic framework, based on the researched ontology and operational process, is presented. The time-series data is stored on an "automated" database server, custom-configured to routinely "collect", "clean" data using predictive analytics to maintain data integrity, store and present the data securely for research. Several techniques are discussed in the paper to fuse different parametric measures using logic and analysis "at the edge", to integrate all compute hardware or cloud capability, using OTA (over the air) techniques. The data collected presents researchers with a rich resource from which to extract, amongst other insights, traffic patterns, weather relationships, risk management dependencies, health risk considerations, time sync error issues that are central to the integrity of any time-sensitive operation. This paper specifically investigates the technology, the process, techniques such as business process workflows, semantic modelling, data analysis, optimisation of tasks, the effect of transmission delays on the results of analysis, predictive techniques, ontegration via APIs and the concept of the "digital twin" to study the operational effectiveness of the Smart City IoT ecosystem.