LEADERS LEAGUE: Can you shed light on the early development of EarthLab since its foundation in 2015, and how the company initially evaluated its service offerings in the expanding space industry?

Thomas Friederich: In 2015, Telespazio, a subsidiary of Thales-Leonardo, recognized the untapped potential in private-sector Earth observation. Consequently, the company established EarthLabs, a network of specialized facilities focusing on various sectors. Key parts include EarthLab Aquitaine, concentrating on viticulture, and EarthLab Luxembourg, dedicated to risk management for insurance and industry. Telespazio France collaborated with Luxembourg-based organizations to create EarthLab as a joint venture, combining expertise in Earth observation, space domain management, and Big Data infrastructure.

The challenges surrounding intellectual property and IP assets related to artificial intelligence involve ownership and how it is valued.

In the current economic landscape, how does your company navigate the intersection between the space industry and insurance and reinsurance sectors, given your technology-driven approach and reliance on satellite data?

Initially focused on New Space, EarthLab has branched out into specialization in artificial intelligence (AI) for automating data-scientists’ work. Its main market being insurance and reinsurance, EarthLab also serves the sector in an Industry 4.0 context, making its positioning difficult to define. The company’s intellectual property (IP) assets are primarily based on the Max-ICS platform, which processes massive data, creates data processing chains, and develops AI models.

The challenges surrounding intellectual property and IP assets related to artificial intelligence involve ownership and how it is valued. Contractual agreements are established at the beginning of a project to clarify the origin and ownership of IP. However, issues arise when discussing the reuse of AI models, which pose significant challenges for companies working in the field of artificial intelligence.

How does your risk management solution apply to the insurance industry and which other sectors is Earthlab targeting?

Our solution provides fast and relevant information in the aftermath of major natural disasters, which is especially valuable for reinsurance and insurance companies. Our mission is to use Earth observation data and advanced artificial intelligence models, such as advanced generative networks (GANs), to estimate the impact of disasters on homes, particularly in terms of destruction.

By comparing images before and after disasters, we can quickly provide high-value-added information that can help insurers proactively manage relationships with policyholders and mobilize funds to compensate them. Despite the opportunities in risk assessment and management, there are also risks that need to be considered. One of the main risks is the accuracy and anticipation of data, as well as the need to adapt to climate change and the increasing frequency of natural disasters. Therefore, it is important to remain vigilant and continue to improve our models and methods for assessing and managing risk.

Ultimately, the objective of the Deep Learning Factory is to simplify the process and accelerate the adoption of artificial intelligence.

What is the Deep Learning Factory?

The Deep Learning Factory is an in-house project aimed at producing deep learning at scale to facilitate and speed up the implementation of artificial intelligence for data scientists. Its purpose is to enable quick and easy experimentation on specific themes or issues without the need for significant investment. It allows for the rapid creation of numerous models, identifying signals and determining whether further investment of time and resources is warranted. Ultimately, the objective of the Deep Learning Factory is to simplify the process and accelerate the adoption of artificial intelligence.

The space sector is relatively unregulated compared to fields such as life sciences. Is this likely to change?

Actually, there are already a substantial number of regulations governing the space sector, which is a broad term that encompasses numerous activities and areas, such as telecommunications, Earth observation and space resources. Each of these areas has its own specific rules.

To give you an example, in my field of activity, Earth observation, we are quickly confronted with defense issues. As you can imagine, this activity is subject to very particular and sensitive European regulatory frameworks, especially regarding the exploitation of Earth observation data. These challenges are even more significant when they are related to artificial intelligence.

Could you tell us what the average Earthlab employee do?

Within our company, the workforce can vary, but on average, we have about ten employees. These individuals primarily have backgrounds in IT and data science. The IT experts mainly work on the development of our platform, while the data scientists engage in highly specific projects for clients from various sectors. These sectors include insurance, agriculture, environment, and sometimes even defense. Our organization, therefore, relies on these two pillars: on one hand, the development and maintenance of our platform, and on the other, the involvement of our data scientists in more targeted projects.

Luxembourg actively supports and encourages the space industry through financial and technical assistance, and serves as a platform for exchange and networking among players.

What advantages does being based in Luxembourg afford for companies engaged in the space sector?

Luxembourg offers several advantages to companies in the space industry. The country actively supports and encourages the space industry through financial and technical assistance, and serves as a platform for exchange and networking among players in the sector. This facilitates the establishment of partnerships and collaborations. Moreover, Luxembourg offers funds to finance projects and sustain growth, and its small size makes it an ideal testing ground for entrepreneurs with innovative ideas. Access to the market and decision-makers is straightforward, and the multicultural population is open to innovation.

Finally, having international shareholders allows companies to benefit from their support in various countries, enabling them to operate internationally with active connections in each of these countries.

There is a misunderstanding among the public and private sectors about artificial intelligence, including technological, regulatory, and ethical aspects.

What are the advantages for businesses of using satellite data?

Our solution addresses not only risk management but also business intelligence and economic intelligence. For instance, satellite data can help anticipate the production of raw materials, which is useful for risk management and economic intelligence. However, there is a lag in technology adoption across all sectors, especially in Europe, compared to the US and Asia. Rather than blocking technologies, it would be more advantageous to create regulated European actors capable of meeting these needs. There is also a misunderstanding among the public and private sectors and the population about artificial intelligence, including technological, regulatory, and ethical aspects.

Aude Ghespière