• Posted 14-Sep-2021

Design of experiments made faster and easier than ever before with AI

New AI platform for designing experiments shortens the go-to-market process for industrial innovation

The Latvian start-up Exponential Technologies Ltd (xT) has developed an artificial intelligence-driven Design of Experiments (DoE) software, called xT SAAM, to help customers succeed in their experiments and data analysis faster and easier than ever before.

In additive manufacturing, material properties can be altered substantially by changing the printer parameters. This is one of the vast benefits of additive manufacturing: material properties can be adjusted to fit specific use cases during manufacturing. These adjustable material properties, together with the design freedom inherent to additive manufacturing, enables the manufacturer to develop new, highly efficient designs, lightweight and mechanically robust structures for space, aviation and mobility applications and much more. On the other hand, this also makes the development of new materials and printer parameters extremely time and cost-intensive.

xT has already collaborated with larger companies in several industrial sectors, e.g. additive manufacturing, chemicals and biotechnology. One of these is Spanish household chemicals company SPB.

Exponential Technologies was founded in 2019 by Pavels Cacivkins, Matthias Kaiser and Girts Smelters. The company has developed, in close cooperation with academic and industrial partners, and industrial data farming platform to help researchers, engineers and data scientists in the development and optimisation of new materials, chemicals, machine and process parameters. The software platform is also used in the management and mitigation of production anomalies.

Fast learning curve for the user

Currently, the most used approach to find new chemical formulations and processing parameters is the use of classical DoE software, often in combination with statisticians, data scientists or DoE experts. DoE software is complicated and requires expert knowledge in statistics.

Consequently, researchers have to master also statistics, next to their domain expertise, or they must inquire support from the limited number of available DoE specialists. This leads to bottlenecks in the R&D process. Furthermore, most SMEs won’t be able to afford DoE experts, which forces them to use less efficient research methods like one-factor-at-a-time (OFAT) or grid optimisation.

xT SAAM automates the process by combining classical DoE methods with novel AI algorithms. This allows domain experts and laboratory personnel to run experiments without the need for DoE expertise in a fast and efficient manner. No statistical knowledge is required from the user. The software acts like a navigation system in which the researcher sets up a destination, and the platform helps find the shortest path.

Consequently, laboratory personnel and data scientists can concentrate on high-level data analysis using the xT SAAM integrated data science toolset, which has a user-friendly interface. The increased R&D efficiency allows R&D managers to lead more teams and projects. The integrated cooperative workspaces further increase cooperation and communication for all involved stakeholders and will enable the storage of all relevant R&D data in one place.

Reducing R&D time and costs

Compared to other techniques, xT SAAM finds satisfactory results with fewer samples required, which additionally reduces R&D time and cost.

The software program xT SAAM helps to find machine- and process parameters and material compositions and many other parameter types without the need for advanced statistical knowledge. The software can be deployed in a wide range of industrial applications, like chemical manufacturing, additive manufacturing, biotechnology and many more.

Exponential Technologies (xT) received EIT RawMaterials Booster funding in 2020, and the start-up is currently developing product testing with customers within the EIT RawMaterials network.

Source: EIT Raw Materials I News (https://bit.ly/3jWdYDo)