Case study · Infringement analysis

KUKA logo
KUKA logo

×

Stilta logo
Stilta logo

September 2026

September 2026

Stilta re-run of a prior infringement case

KUKA’s IP team had already worked this infringement case to a clear, defensible conclusion. In evaluating Stilta, KUKA put Stilta to the same task. Grounded in agentic evidence collection, Stilta landed where the case conclusion did in a single autonomous run, with no human input and no prompting, in about 20 minutes.

KUKA’s IP team had already worked this infringement case to a clear, defensible conclusion. In evaluating Stilta, KUKA put Stilta to the same task. Grounded in agentic evidence collection, Stilta landed where the case conclusion did in a single autonomous run, with no human input and no prompting, in about 20 minutes.

KUKA’s IP team had already worked this infringement case to a clear, defensible conclusion. In evaluating Stilta, KUKA put Stilta to the same task. Grounded in agentic evidence collection, Stilta landed where the case conclusion did in a single autonomous run, with no human input and no prompting, in about 20 minutes.

The situation

KUKA’s IP team works at the forefront of its field, actively bringing new technology into high-stakes patent work. As part of this work, the team selected a real infringement case that it had already assessed internally. This gave KUKA a known conclusion against which Stilta could be compared.

The case concerned a patent for logistics technology. A question had surfaced from the field: signals from KUKA’s own developers and from trade fairs suggested a competitor’s product line might be using the protected technology. The KUKA IP team looked into it by gathering the public evidence, reviewing the competitor’s product web pages, discussing with the developers and project managers, and mapping the evidence against the patent’s claim limitations. The conclusion: no infringement. The competitor’s products didn’t carry all of the features that define the patent.

The case concerned a patent for logistics technology. A question had surfaced from the field: signals from KUKA’s own developers and from trade fairs suggested a competitor’s product line might be using the protected technology. The KUKA IP team looked into it by gathering the public evidence, reviewing the competitor’s product web pages, discussing with the developers and project managers, and mapping the evidence against the patent’s claim limitations. The conclusion: no infringement. The competitor’s products didn’t carry all of the features that define the patent.

The case concerned a patent for logistics technology. A question had surfaced from the field: signals from KUKA’s own developers and from trade fairs suggested a competitor’s product line might be using the protected technology. The KUKA IP team looked into it by gathering the public evidence, reviewing the competitor’s product web pages, discussing with the developers and project managers, and mapping the evidence against the patent’s claim limitations. The conclusion: no infringement. The competitor’s products didn’t carry all of the features that define the patent.

Using Stilta

KUKA then re-ran the same patent through Stilta. Using public evidence only, Stilta reached the same conclusion, no infringement, in about 20 minutes.

It got there in the way a practitioner would: it first broke the patent down into its claim limitations, then searched out the competitor’s relevant products and assembled the supporting evidence, and finally mapped that evidence to each limitation in a claim chart, making clear where the products matched and, decisively, where the defining features were absent.

The objective, methodological approach signals restraint and is built on first principles, which is what top-tier IP teams like KUKA’s value most.

The outcome

~20min
single end-to-end autonomous run
500+references
scanned across public product sources
85–90%
less time than the manual assessment

KUKA’s prior assessment

Stilta: autonomous agentic run

KUKA

Products identified by manually reviewing the competitor’s product web pages

STILTA

Same key products found by Stilta agents searching across the web and public product sources

KUKA

Key evidence from public product and web descriptions, gathered and charted by the team

STILTA

Same evidence found by Stilta agents scanning the competitor’s product pages and documentation across 500+ references

KUKA

Evidence-to-claim mapping, charted limitation by limitation, highlighting the gaps

STILTA

Same gaps identified. Agents break down and analyse each limitation, using the latest AI models

KUKA

Verdict: no infringement

STILTA

No infringement

KUKA

Effort: assessment of the patent, mapping the claims, gathering evidence, meetings with developers and project managers, final evaluation

STILTA

~20 min, 0 human inputs, single end-to-end agentic run using only the patent id

This was Stilta’s autonomous run on its own, with no attorney in the loop. Stilta does the heavy lifting: the searching, the evidence-gathering and the mapping. In practice an expert then builds on the run in attorney-led work.

“We gave Stilta a completed infringement case without sharing our analysis or conclusion. It independently found the relevant products and public evidence, identified the same decisive gaps, and reached the same conclusion in 10–15% of the time.”

Stefan Schmid, KUKA IP team

Get started with Stilta

Evidence-first patent analysis that builds the strongest case. Book a demo to see it on your own work.

Book a Demo