En-ROADS is specifically designed to be fast enough to see the results of your choices in real-time.
There are two parts of “how” to make En-ROADS so fast: 1) model design choices, and 2) technical computer science choices.
1. Model design choices
Our modelers continuously assess what details need to be a part of the En-ROADS model in order to provide useful, accurate insights about climate solutions. The team adds geographic, technical, economic, and climate physics detail when it is necessary to answer questions in a rigorous manner. This strict modeling design policy is critical to making a fast-running model because, by limiting what is in the model, only the mathematical equations that are necessary for insights are being calculated by the computer.
2. Technical computer science choices
With only the mathematical calculations necessary in the model, the next step is making the model fast enough to use over the internet. Climate Interactive has a long history of deploying decision support tools and models, and we’ve that learned relying on internet connections and the cloud does not produce a good user experience.
The En-ROADS simulator runs on your computer when you visit our website. When you load the webpage, the simulator code is loaded into your browser and then no server connection is needed. In fact, you could disconnect from the internet (e.g., turn off wifi) and still run the model – just don’t close or reload that browser tab. The simulator runs all of the En-ROADS equations every time you move a slider.
To help publish models, Climate Interactive developed an open-source programming toolchain called SDEverywhere. This tools converts our modelers’ work from Vensim – a system dynamics modeling tool – to the C programming language. The SDEverywhere toolchain then uses Emscripten to transpile the C into WebAssembly, which works in your internet browser. WebAssembly can run equations very quickly, and En-ROADS can run all of its thousands of equations in about 30 milliseconds.