Simanfor: Artificial intelligence meets nature

SNGULAR participated in an integral update of the Simanfor platform, which runs pioneering forest simulation models.

Predictive models are increasingly being used in sectors as diverse as finance, eCommerce, insurance, telecommunications or pharmaceuticals. The boom means that the models are evolving quickly and are becoming a critical tool for sectors across the board.

But what exactly is predictive modeling? It’s a mathematical process that seeks to anticipate future events or outcomes by analyzing data to find patterns and trends. Thanks to modeling statistics, machine learning and data mining, it is increasingly accurate at predicting the future and what impact new actions can have.

The University of Valladolid in Spain launched Simanfor, a system that runs forest simulation models, to take advantage of the power of predictive modeling for forest management. It allows researchers to manage forest inventories, project the dynamics of the forest system through empirical models and develop and evaluate user-designed forest regimes. It is the first-ever forest simulation to be based fully in the cloud.

«I’ve loved being able to work on the entire life cycle of this technology project that is directly applied to the environment»

Jorge Prudencio.

Our colleague Jorge Prudencio, a cognitive solutions architect, expressed his passion for the project. “Seeing something that took so much work become useful and productive is extremely satisfying,” he said. Simanfor is now being used both to train new forest engineers and ensure the sustainability of the forest through better management.

Update and redesign of the Simanfor platform

A SNGULAR team consisting of specialists in machine learning, data, cloud and frontend was involved in completing an integral update of the platform.

The first step was updating the entire internal simulation motor, which allows researchers to estimate the future effects of regimes that they are considering. It uses current tree inventories and can simulate applied actions (cutting down trees, change over time, etc.).

At the same time, they completely redesigned the web interface and enabled a Rest API to allow for scalability and the development of new interfaces in the future. With these changes, Simanfor was made more user-friendly. The update also made it possible for the platform to become available on more devices immediately. For the front and backend development of the project, they used Angular, NodeJS and MongoDB.

The goals of the initiative, which at the beginning, aimed to update the simulation motor for a completely new one, were later expanded when the need emerged to also build a web interface to facilitate the use of the new motor.

The team that led the project started from a base of the many of the ideas and functionalities that Simanfor had already incorporated in its previous version and added several technologies and functionalities that were more efficient and evolved.

Following a scrum work model and staying in constant contact with the University of Valladolid, Jorge Prudencio said that each control point with the university was invaluable. At each point, the SNGULAR team presented the results through a product demo and collected critical feedback to adapt the platform to the client's needs.

Felipe Bravo, professor at the University of Valladolid and director of the University Institute for Research in Sustainable Forest Management, says that thanks to this collective effort "Simanfor has positioned the University of Valladolid as a leading reference in forestry research."

"The collaboration with SNGULAR exceeded our expectations regarding the development of Simanfor and the incorporation of valuable new tools that were the result of multidisciplinary work," Bravo continued.