Systems Engineering: From Prototype to High-Performance Solution

It was developed by an external software development company. When Systems Engineering A/S was just starting, they used paper, then Excel, to help other companies design engineering systems. Both tools had similar drawbacks: they were too time-consuming and not scalable enough. The resulting application prototype, with Ruby on the backend, fulfilled the business requirements but turned out to be not scalable enough. It was possible to design an engineering system with it but it couldn’t handle the magnitude required by the clients (tens of thousands of elements possible while billions were needed).

The goal we started with was: to rewrite the existing prototype for efficiency and scalability. To handle such immense numbers of elements, the new app was supposed to use a more efficient Neo4J graph database, and Vue.js on the frontend.

The mobile app allows users to quickly pull up key information from the web app, including details about the elements that underpin a system’s design. The interface of the mobile app needed a redesign to improve User Experience. The technology that it was using also required changing for security and future extendibility reasons (from React Native JavaScript to React Native TypeScript).

Although the core functionality was to stay the same, some new features were needed (like the ability to switch between workspaces), as well as an integration with a new API that the main web service had.

As a next step, we’ve collaborated with Systems Engineering to break the requirements and ideas into smaller, tangible tasks.

We’ve repeated this process iteratively throughout the web prototype development and the mobile app refactoring project. Working in this way, we’ve carried out a number of technical and functional improvements such as implementing end-to-end tests or figuring out the best way to add functionality where users can join workspaces.

To rewrite and scale the PoC, we started from scratch by creating an architecture for the Neo4J backend database, and building the frontend part. The PoC redevelopment lasted 2 months. Then we proceeded to build new features.

We built two core modules: one that deals with designing electrical systems, and another that allows combining this aspect with two other elements of an entire engineering system. We also created a module that enables comparison of two similar versions of an engineering component and accepts or rejects the differences. We also designed and developed a version of the app for individual, non-corporate engineers.

• Regularly discussing priorities and plans between Systems Engineering stakeholders and developers.

• Translated the requirements into technical solutions and implemented an iterative approach to drafting solutions.

• Set up a self-organizing team of developers who proactively seek to understand the business context.

Although the underlying web service has changed significantly, the mobile app stayed functional and continued to facilitate communication around the systems being designed.

The workspace functionality that we’ve implemented allows the users to have different viewpoints and discussion angles.

We were able to deliver software that handles billions of elements in engineering systems while staying scalable and performant. One of the updates we included improved the speed of the app by 10-20 times.

We've also created a mobile companion app for the main software.

• Finding creative technical solutions for efficiently displaying large numbers of elements.

• Actively looking to understand the business context and proposing solutions to address the client's needs.

• Iteratively working with Systems Engineering to find the best solutions possible for the mobile app.