How DeLaval use IPRally to find prior art that conventional methods miss
A new tool for new technologies
DeLaval has been in the milking equipment industry for almost 140 years and is one of the leading innovators in its category. Dairy technologies have constantly developed over the years, and sensors, algorithms and internet of things are now a natural part of the once purely mechanical field. In short, the competitive landscape has broadened vastly. When Oskar Christensen started working as a patent engineer at DeLaval, he knew that it was impossible to be an expert in all these technology fields. But since his previous role as a patent examiner, he was familiar with IPRally, and knew it made technical expertise less critical in the search process.
“Out of all the AI solutions I tested at the PTO, I thought IPRally worked the best. So when I joined DeLaval, I wanted to get us started with IPRally. I knew its strengths and I knew that it would be very quick to learn for my new colleagues – you can basically just start searching.”
A different search angle
Oskar uses IPRally primarily for invalidity searches. A big challenge, he says, is searching in technical fields you are not familiar with. Finding documents that the patent examiners might have missed, for example when working on an opposition case, is also tricky. But with IPRally's built-in technical understanding and natural language processing, a new layer of hidden prior art opens up.
“Trying to beat a patent examiner is a huge challenge. They are skilled professionals who are experts in their technical fields. It's very unlikely that you will find something they have missed if you use the same methods, classes and boolean operators that they have used. So your best bet is to try a different angle. It can be searching for documents that are misclassified, poorly translated or that use an odd terminology, and IPRally is great at finding this kind of hidden prior art.”
Less noise, easier review
Another challenge that Oskar mentions is the constantly growing body of patent data. IPRally, he says, often provides a good start with less noise.
"If you get 5000 documents in the results, you don't even want to start looking at them. But with IPRally there's less noise. And with the automatic highlighting, you can quickly see what's interesting and what's not. I almost always start my searches in IPRally, especially if I don't know where to start. I might use our conventional search platform later on if I need to dig deeper, but many times a quick search in IPRally is sufficient."
According to Oskar, IPRally brings both speed and clarity to the review process.
"You see how the AI works and why the hits were found, and you can nudge it by modifying the graph. So it makes you feel more in control than with other tools. The most relevant documents often come at the top picks in the search results. So if I limit myself to 50 documents or something like that, I am pretty confident that it's a good search", says Oskar Christensen.
