Aledia Patent Portfolio Analysis

One pulse, arriving eight years late

The Series D was announced on 1 October 2020: $94.1M, a single event in a record that had been accumulating for eight years by then. The first family was filed on 18 September 2012. Seven followed that year, seventeen the year after. By the time the money arrived, 2019 had already been the heaviest filing year on the books, with twenty five families. The capital did not start the work. It landed on top of it.

Read the years in order and the pulse looks less like a beginning than a confirmation. Twenty four families in 2018, twenty five in 2019, twenty in 2020. The rate barely moves across the raise. Whatever the round paid for, it was not a change of subject, and the filings kept coming through 2021, 2022 and 2023 at roughly twenty families a year before easing to five in 2024.

Most of this has already been examined

Two hundred and eleven families in total. One hundred and eighty two granted. Twenty nine published and still pending. Eighty six percent through examination and allowed. For a company that started five new families in 2024 and has a published filing dated 30 May 2025, that ratio leans hard in one direction: almost all of this record has already passed in front of an examiner somewhere.

A pending stack is a claim somebody has made. A granted stack is a claim somebody else has checked. Records built fast usually sit the other way round, with the weight in front of examination rather than behind it. Here the weight is behind. The twenty nine unexamined families are the newest thinking. The other one hundred and eighty two are settled ground, and they were mostly settled before 2020.

A pending stack is a claim somebody has made. A granted stack is one somebody else has checked.

One subject, and a few close neighbours

H10H, light-emitting devices, carries 161 of the 211 families. H01L, semiconductor devices, carries 140. Families can sit in more than one group, so the two overlap heavily rather than competing. Then H10W at eighty one, H10P at thirty nine, and two equal tails of twenty nine: G09G and H10F. The shape is a single subject with a small set of adjacent areas, not a spread of separate interests.

The titles agree with the classification. The recent published filings name an optoelectronic device and manufacturing method, a 3D light-emitting diode and associated manufacturing method, a display pixel comprising light-emitting sources and a display screen having such display pixels, and a method for manufacturing at least one optoelectronic device. Manufacturing method keeps reappearing. Thirteen years in, the claims still return to how the thing is made.

Read most often by people elsewhere

Commissariat Energie Atomique cites these filings 130 times, more than any other organisation. Then Osram Opto Semiconductors at seventy two, Lumileds at sixty five, and Samsung Electronics and Valeo Vision at thirty seven each. The record says nothing about why any of them cite. It says only that they do, and that the citations run back to filings that begin in 2012.

Citation counts are a crude instrument, but concentration in them is readable. One hundred and thirty references from a single organisation is not incidental. It is repeated return to the same body of work while later applications were being drafted or examined. Four of the five heaviest citing organisations are based outside France. This work is being read in places where it was not written.

Eight offices, two of them deliberate

Eight offices, with PCT set aside: the United States, the European Patent Office, China, Japan, Korea, France, Brazil and Taiwan. The first six are what you file when you want to be readable in the major manufacturing jurisdictions. Brazil and Taiwan are choices. Each one costs fees, translation and years of prosecution, and neither is a box anybody ticks by default.

France is the home office and the origin of the September 2012 filing. The rest were decided family by family across thirteen years. Put that next to 182 grants and you have a company that paid for examination in places it was not obliged to enter. The Series D was $94.1M. A large share of what it was funding had already been filed, prosecuted and granted.

The order of events is the argument

Here is what the file shows. A first filing in September 2012. Two hundred and eleven families and 201 named inventors, three of whom, Ivan-Christophe Robin, Frédéric Mercier and Philippe Gilet, appear on 46, 41 and 34 families respectively. Eighty six percent granted. Eight offices. The heaviest filing year was 2019, the year before the round. The money came last, not first.

Samsung Display, Jiangxi MTC Semiconductor, LG Display, Samsung Electronics, Nichia and ams OSRAM all file in the same CPC groups, and all of them at far greater volume over the last five years. Aledia's position was never going to be volume. It is thirteen years of examined grants in one subject, carried by a small recurring group of names, and it was substantially in place before any of the Series D was spent.

The capital arrived after the examination, not before it.