The IFI Interview: Donata Gabelloni, Head of Research and Development at Erre Quadro

Numerous successful companies have their roots in academia. Count research company Erre Quadro among them. Erre Quadro, which develops cognitive engines for the study of technical documents, arose from the work of three Italian researchers at the University of Pisa, who wanted to apply engineering design methodologies to technical text analysis. Because few texts are more technical than patents, these University of Pisa academicians leaned into these documents as one of their first areas to apply natural language processing. That was in 2008. Today, the company has a suite of products applying AI to the understanding of all sorts of highly technical documents.

Donata Gabelloni, who leads R&D for Erre Quadro, joined the company in 2014 as a junior patent analyst, after earning her PhD in Rome. (She studied at the University of Pisa for her bachelor’s and master’s degrees). Since then, much of her work at the company has been steeped in the intricacies of patents.

IFI CLAIMS talked with Gabelloni in February, on the heels of Erre Quadro’s announcement of opening a new office (and a new market!) in Denmark. She talked about the particularity of patent language, what industrial design theory brings to the understanding of patents, what the company is doing with patent images to increase comprehension of these often obscure documents, and why Erre Quadro chose IFI CLAIMS data as the mainstay of the company’s patent analysis. Below is an edited version of our conversation.

IFI CLAIMS: Tell us about your background.

Gabelloni: I started in industrial management engineering at the University of Pisa where I earned my bachelor’s and my master’s degree. Then I did a PhD in industrial management engineering at University of Rome Tor Vergata, while still collaborating with the research group in Pisa. During my PhD, I worked on functional analysis, a design methodology that is used to analyze a product, such as identifying the main functions that the product carries. Functional analysis is used in different phases of the new product design process. For example, functional analysis is the base of several other design methodologies or quality function deployments. I also worked on crowdsourcing at this time. Then, in 2014, I started to work at Erre Quadro as a junior patent analyst. That was the first time I worked with patents. As time went on, I filled different positions at Erre Quadro. I managed the operation. I worked on the proposal for a European-funded project. I also worked on the economic and financial parts of the company. Then I started to be involved in the research activities of the business—in particular, in the design and logic of the algorithms to analyze technical data. Currently, I coordinate internal research projects, especially research projects on patents.

IFI CLAIMS: You said that you started to work with patents when you first took the job at Erre Quadro. Was that difficult in the beginning? Or was it a natural progression of your studies?

Gabelloni: Patents are very particular documents. They are technical documents, but the language of the patents is not just technical. There are also legal aspects. So at first, it took me a little time to understand the new language, but my background in functional analysis helped me to understand how to perform the queries and how to use the tools to do the patent analysis. It was a new world, but not so distant from the one I was already in.

IFI CLAIMS: Tell us about the company, how it started, how it developed, and then eventually how patents came to find their place there.

Gabelloni: Erre Quadro was founded in 2008 by three researchers from academia—from the University of Pisa, in fact. They tried to apply the engineering design methodologies to the work of the technical text analysis to develop software. Patents were one of the first areas they started to work on. Erre Quadro was founded to extract relevant information from the technical text using natural language processing algorithms. It started as a service company. First they developed the software tools and internal search engine to analyze patents and then started to provide service to customers. For example, patentability analysis, freedom to operate, technical landscapes to map technological fields, and insight analysis using patents and other kinds of documents like scientific papers.

Erre Quadro is a spinoff brand from the University of Pisa. The goal was to automate the interpretation and retrieval of information from documents. Then in 2018, the company transitioned into a product-based company. We launched Innovation Reveal, a platform where our customers can extract information from a patent set. This is one of our main products. We started to expand our tools to other technical texts like specification of products. At Erre Quadro, we have a multidisciplinary team—engineers with different mechanical, biomedical, chemical and robotics backgrounds. We also have a group of computer scientists that develop the algorithms and the codes of the tools. Rounding out the team are mathematicians and physicists.

Our market is mainly in Italy—at least at the moment. Last year, we opened a new branch in Denmark, so we’re expanding. Our clients are both small and medium enterprises, but also large companies. They’re from different sectors. For example, we work a lot in the biomedical sector, but also in the energy sector, manufacturing sector and automotive sector. We collaborated with national and international institutions on big technological studies and foresight analysis. For example, we performed a foresight analysis on mechatronics for a region of Italy.

IFI CLAIMS: If an automotive, energy or biomedical customer comes to you, what typical problems do you help them solve? And how do you go about solving it for them?

Gabelloni: Big companies generally ask us for a landscape analysis. For example, if they want to investigate a new technological area, they ask us for some sort of technical review. And generally they are interested in our platform, Innovation Reveal, because they have a big patent portfolio to manage and they want to monitor their competitors. On Innovation Reveal, we provide benchmarking analysis on patents. Sometimes the client has a problem with the specific patent or patents of a competitor, so they ask us for our analysis, such as an infringement analysis, because of our accuracy and high quality. They also look to us for patentability analysis and freedom to operate. A company might want to go into another market, for example, maybe in the U.S. where it’s mandatory in some biomedical fields to have the freedom to operate. So our clients also ask us to help them understand those things.

IFI CLAIMS: What typical public sector problems are you helping to solve?

Gabelloni: With the public sector, generally we perform foresight analysis. Public sector clients want to see technology trends in order to understand which areas to invest public funds in. In the past, we have worked for the Tuscan region. We have also done a study on biometrics technologies for the European agency, Frontex.

IFI CLAIMS: What else is important to understand about Erre Quadro?

Gabelloni: One of our strengths is our expertise and experience in engineering design theory. We also understand the environment and requirements of the patent analyst, their needs and difficulties in evaluating patents. So our knowledge of patents is an advantage when developing the tools to analyze those patents because we know the structure and language of the documents. We have also developed a solid knowledge base and the technical ontologies that we use in our machines, which allows us to perform the research with a high degree of accuracy and precision.

IFI CLAIMS: What are some overall trends you’ve noticed with patents? Any future predictions?

Gabelloni: Our market is currently in Italy, so we see just part of the picture. But what we’ve noticed is a strong trend in patenting green products. A lot of clients ask for studies on green products. Batteries, for instance. Actually, companies started asking us for work on batteries eight years ago, but these requests are definitely increasing. The other thing they ask us for is analysis for solutions to reduce energy consumption or solutions that use eco-friendly materials. Another trend we’re seeing is patents on software or methods related to software. Software is not easy to patent in Europe, which is a big issue that needs solving. Software is so important now in many industrial sector applications in addition to daily life, so companies are trying to understand how to protect their software. It’s very complex. Companies want to know if there are patents on solutions that are similar to solutions they are developing.

IFI CLAIMS: To avoid infringing? Or for their own competitive understanding?

Gabelloni: Both. We have more and more companies asking us for freedom to operate because they’re looking for financing from venture capitalists. So they need an evaluation of the solidity of the assets that are protected. By the same token, we see an increase in requests from venture capitalists who ask us to look at the patent portfolio of a target company they are evaluating for investments. They are using us for the technical part of their investment due diligence.

IFI CLAIMS: Why did you become a customer of IFI?

Gabelloni: We have an internal search engine that we use for our analysis. Good quality data is crucial, which is why we chose IFI. IFI provides us with complete access to the major jurisdictions. Good patent coverage is important to us because our customers want an overview from all around the world—or, at the very least, the major jurisdictions. Another important thing is that with IFI data, we have direct access to the database—a relational database with entities. We can import that data onto our machine. Because of our internal tools, this is important for us. Another advantage of IFI patent data is that there are daily patent updates, which means our database is updated daily. That’s another very important aspect because our clients want to know that our patentability analysis has the most recent data. Access to patent images through the API is a big thing to us. Finally, I would say that not only do we get the classic metadata of the patents, text and images, but IFI adds information—metadata like the assignees’ country code, and the parent company of the assignee. This is useful for us to perform analysis and gather statistics for our customers.

IFI CLAIMS: We’re curious. Why are patent images so important to your company?

Gabelloni: The reason why we need the images is because we have a specific tool to analyze those images called “TagMyPic.” We can tag the images, find the name of the components in the text, and then link them directly to the images. It accelerates the reading and understanding of the patents. It’s a tool that is important to our internal patent analysts and also to the clients who use it because it speeds up their understanding of a patent too.