In my previous article, I introduced the concept of Digital Twin. The state of a Digital Twin (DT) relates to data gathered from sensors and external systems. That setup ensures automated, up-to-date digital representation of monitored entities. Today, I would like to focus on the Data Structure aspect of DT. Designing that Data Structure is an iterative, manual process. We extend the Data Structure to improve the functionalities provided by DT.

Digital Twin is a virtual representation of a physical device. It consists of Data Structure, Logic, and Visualization (in my opinion, an optional part of the Digital Twin). We use sensors to gather data about the monitored device and store obtained information using the Data Structures of Digital Twin. We can get additional information about the device from External Systems and also store it using the Digital Twin’s Data Structure.

Proof of Concept (PoC) is a very misleading approach, and I strongly recommend avoiding it. Let me explain. PoC makes the project team focus on the technology and verification of the technical feasibility of some solution. That mindset is one of the main reasons for failure in the IoT domain because: We lose the objective (technology is not the end goal of any IoT deployment). We lose time (contemplating various technical solutions can take infinite time without producing any outcomes).

What is the ultimate advice I can offer as the Internet of Things Advisor and Trainer? Do not deploy an IoT system… … because everyone is doing so, … because that is the way to stay relevant, … because people say that you will fall behind. A poorly designed and implemented IoT solution… … won’t modernize your enterprise, … will cost you time and money, … will harm your operations.

Introduction I received an interesting question on the DevHeads community discord: “When to use one over the other between Amazon DynamoDB and S3 for storage?”. Short answer Two principal factors define data storage: The nature of the data itself. The way we want to consume it. Discussion For our discussion, I will categorize the data provided by IoT devices into four categories: Time-series telemetry readings. Complex readings. Batch data. Custom. By no means is that a comprehensive list, but it illustrates the primary use cases.

I tried to make ChatGPT generate the AWS Cloud infrastructure required to send telemetry messages from devices to the cloud backend. I asked it to use the AWS CDK to describe infrastructure as code so I could deploy that setup and test if it actually works. That is the solution I wanted to achieve: AWS Cloud Backend If you are curious about the outcome, please watch the video below. I also share my conversation with ChatGPT so you can review it and try to deploy in your AWS account.