GreenTropism
Delivering AI-Powered Spectral Analysis Model

Over the years, spectroscopy systems have been a reliable and robust supplement to conventional titration methods while monitoring industrial manufacturing processes. However, the heterogeneity of the spectrometer’s hardware makes it difficult to use. Besides, due to the complex nature of the system, operators find it impossible to use various spectrometers (lab, process, miniaturized) as a network of sensors. In such a situation, what organizations need is a new generation of simple spectrometers that can be used for diverse purposes and deliver results under any circumstances.

Established in 2014, GreenTropism is a promising French startup specialized in embedded software catering to industrial spectroscopy. The company has developed a groundbreaking AI-powered Spectral Analysis Model for managing and processing spectral and related data. “By creating robust algorithms for normalization and simulating spectrometers and AI, we can manage a fleet of spectrometers, and process and merge data under various formats,” says Anthony Boulanger, CEO of GreenTropism. With such unparalleled capabilities, GreenTropism offers its solution to different markets, including medtech, agri-food, industrial processes, and environmental industry

The driving force behind GreenTropism’s embedded software is its data analysis algorithm, which is dedicated to applications in material control or sorting, contamination control, and many more. “Our algorithms have the capabilities to function in a cloud-based environment according to customers’ needs,” states Boulanger. The company’s AI software is extremely fast and accurate when it comes to model development and predictive analysis of the collected data.

As it is dedicated to spectral data, it can be much more efficient than general AI. This specialization is not limited to one specific industry; as spectral data is present in many technologies (UV, Visible, NIR, Raman, LIBS, etc.), the company’s AI software can be used in various markets and applications.

GreenTropism’s solutions rely mainly on three components: GT-DataManager, dedicated to the spectrum’s exploitation, KAÏSSA to build predictive models, and GT-Lib to exploit models. GT-DataManager is used in the laboratory to develop experimental plans, a nomenclature for samples and spectrometers. KAÏSSA is an AI tool dedicated to creating chemometrics and spectral analysis models. It chooses and combines the best spectral pretreatment algorithms and selects different modelization strategies to provide accurate and reliable predictive models within a few minutes. GT-Lib gives easy access to predictive models built by KAÏSSA. It can be embedded in smartphones or computers and comes with a user-friendly graphical interface.

GreenTropism also banks on several partnerships with research institutes such as INRAE and a team of biologists and data scientists specialized in bioinformatics, chemometrics, predictive analysis, and machine learning algorithms. The company is also supported by renowned investors and works with multinational firms. For instance, GreenTropism provides its AI software in white-label mode to allow a major spectrometer manufacturer to deploy a spectrometer with AI capabilities for non-expert users. In another instance, GreenTropism worked with a food and beverage company (UK, France, and Germany-based) to deploy software components, which helped them in managing a global database with automatic model updates based on the spectrometers of each factory.

The future holds even brighter prospects for GreenTropism. The company plans to release its software suite on Android and iOS platforms which will allow spectrometers to communicate via Bluetooth. Users can now access and manage spectrometers individually or as a fleet via easy-to-use smartphone applications and in remote mode. “Our ultimate aim is to grow and manage European projects from our Paris offices, and to open in the U.S. and Asia in the coming quarters,” concludes Boulanger.