, , , ,

Pre-Collision Warning for Patent Applications

pc_a

How to Reduce Patent Prosecution Risks

Among the many automotive industry advancements that have improved driver safety, some of the most notable are the inventions that use predictive modeling and analysis to reduce accidents. Audi’s Pre-Sense technology is an example, combining hardware, sensors and software to warn drivers of approaching dangers and potential collisions. When identified risks exceed a safety threshold, a host of actions and system responses occur, including acoustic and visual warnings to grab the driver’s attention. It’s a fast, effective and highly automated system of managing risk. 

What does this have to do with our data analytics obsession at ipQuants? 

Our “corporate why” is helping companies make better IP decisions by better managing risks inherent in the many day-to-day decisions that IP teams and their legal counsel must make. Since 2018, we’ve been using mathematical modeling and data analytics to tackle this challenge. During this iterative journey, we’ve discovered irrefutable proof that, much like predictive modeling helps drivers avoid accidents, similarly modeling IP data can help organizations lower the risks of making patent prosecution errors. 

Given the size of many IP portfolios, there’s a lot of room for error. Consider that a company with a portfolio of 4,000 or 5,000 granted patents and hundreds of in-progress applications may easily have several hundred more to prosecute annually. Until recently, it has been almost impossible for applicants and law firms to understand when a patent application enters what we consider “risk territory.” We categorize this as a state where the likelihood of not being granted exceeds the likelihood that it will be. 

Prosecution Risk Modeling

To provide the same degree of objectivity and obviousness that predictive safety systems communicate to vehicle drivers, we developed the Qthena Prosecution Risk Score (Qthena PRS). Ranging from 0 and 100, with 100 being the highest risk score, it’s a metric that quantifies and indicates an objective measure of risk for any pending European and US patent application. 

Qthena Prosecution Risk Score

How does this number aid patent applicants and law firms? Much like the predictive vehicle safety systems that trigger responses after a threshold has been reached, the danger threshold in our system is 75. From analyzing tens of thousands of EPO and USPTO patent applications, we know that when the Qthena PRS is 75 or above, an EP patent application has entered “dangerous territory.” The likelihood that the patent application will be rejected becomes increasingly real. We designed the system so that customers are aware of this very real risk to their application. As with predictive systems, the objective is to capture their attention while they still have time to take corrective action. 

Verifying Qthena PRS Accuracy

Does it work? (We know you’re wondering this). You may be even a little skeptical about the real-world utility of data analytics in IP. Below is a summary of a backtesting study we did across a diverse group of applicants and law firms. We compared the Qthena PRS calculation to EP prosecution data from a variety of industries, including chemistry, IT, aerospace and pharma. Results were similar, regardless of whether they were associated with corporations or counsel. When the Qthena PRS was 75 or higher, those patent applications were increasingly unlikely to be successful.

  • 89% of Microsoft’s EP applications with a Qthena PRS 75 or more eventually “died” 
  • 97% of Siemen’s EP applications with a Qthena PRS 75 or more died 

Takeaway: Qthena PRS was close to 100% accurate in predicting which Siemens patent applications would not be granted. (OK, our predictive modeling wasn’t perfect but perfection is one of our goals and our data scientists will continue to close the gap in the coming years). 

Risk Mitigation When the Data Looks Bad

Without the transparency that access to advanced analytics like Qthena PRS provides, you could very well be in the dark about an at-risk application until it’s far too late to react and implement corrective measures. After all, if you don’t want to look back and relive that nasty “we just lost our patent” shock, you need a danger signal to know to act today. 

How might you respond? While we certainly don’t claim that Qthena PRS will save lives like Audi’s Pre-Sense technology, understanding at-risk applications could very well mean that you have sufficient time to implement countermeasures to save the patent application on behalf of the applicant.

  • You might create a policy that applications handled by junior staff are escalated to senior partner review if their Qthena PRS reaches 75 or above 
  • If the patent application is a core asset of the applicant, you may decide that any measure to protect the application must be considered 

Implementing Qthena Risk Insights in the Real World

We’ve made accessing and understanding patent grant risk exceedingly simple. No patent expertise is required. All the insights can be easily and quickly integrated into any CRM or docketing system, or accessed online through ipQuants’ Qthena platform. Think of it as one-click IP risk management that can be easily added as an agenda item to a monthly or quarterly risk management schedule with legal, R&D and finance teams. And, yes, it does provide an edge to those in the know. 


Image copyright of Audi AG



Below table shows backtesting results for European patent prosecution cases that had a calculated Qthena EP PRS of 75 or above and had a decision date between 2020-01-01 and 2022-01-01.

The results clearly show the high accuracy across a diverse group of applicants, law firms (agent) and industries. For Airbus, for example, 96% accuracy was achieved. Meaning, when the Qthena Prosecution Risk Score (PRS) was 75 or higher, i.e. “high risk territory” was entered, 96% of those high risk Airbus cases were indeed not granted by the patent office. In similar fashion, the Qthena PRS showed high accuracy across the other portfolios as shown below.

EP Party
Qthena Risk Score Accuracy
Party type
Airbus
96%
Applicant
Leica Microsystems
100%
Applicant
Mitscherlich & Partner
89%
Agent
Google
87%
Applicant
Nokia Technologies
89%
Applicant
Plasseraud
94%
Agent
Microsoft
89%
Applicant
Roche Diagnostics
100%
Applicant
Ericsson
91%
Applicant
Siemens
100%
Applicant
Novartis
100%
Applicant
BASF
100%
Applicant
ExxonMobile
100%
Applicant
Novozymes
100%
Applicant
Dräger
100%
Applicant
Uber
92%
Applicant
Facebook (Meta)
98%
Applicant
Meissner Bolte
93%
Agent
Nike
91%
Applicant
Henkel
100%
Applicant
Hewlett Packard
91%
Applicant
Elkington and Fife
95%
Agent
Müller-Boré & Partner
90%
Agent
Michalski Hüttermann
90%
Agent
Potter Clarkson
92%
Agent

Want to see how the Qthena Prosecution Risk Score performs for your organisation? Contact us here and we are happy to show you.


Follow us on LinkedIn and sign up to our Newsletter here. Want a demo? Schedule a time here