Quantum Information

Patents play an important role in the commercial exploitation of emerging technologies. It it therefore not surprising that more and more patents related to quantum information (QI) technologies are granted. D-Wave Systems, the Canada-based company behind the controversial D-Wave Two quantum computer prototype, holds over 260 patents. MagiQ, one of the first companies to supply equipment for quantum cryptographic communication, owns more than 160 patents in that area. Apart from D-Wave Systems and MagiQ, there are many other contestants that heavily invest in QI technologies and who demarcate their innovations by means of patents. In this article I'll try to provide a concise overview of the patent activity in QI technologies. For readers unfamiliar with either patents or QI technologies, I've compiled two primers.

A brief primer on patents: Only technical inventions, methods and implementations are patentable, whereas scientific discoveries, mathematical methods, algorithms or software as such cannot be patented. Further criteria are novelty (no "prior art"), the involvement of an inventive step, industrial applicability and non-disclosure of the invention until the patent is published. Patent applications can be submitted either to national patent offices, or to two international organizations: The EPO (European Patent Office) for up to 40 European countries, and the WIPO (World Intellectual Property Organization) for up to 148 countries. The contents of a patent application become publicly available 18 months after the initial filing date. Hence there is no such thing as a secret patent, and in fact you can google them! Although patents prohibit the commercial exploitation of the invention in the countries for which the patent has been granted, they can be sold or licensed to third parties, and non-commercial usage is always permitted. Patents are valid for up to 20 years, then everyone can freely exploit the invention. Patents are categorized by means of the IPC (International Patent Classification) code scheme, and patents may belong to more than one code. Some patent offices also use the CPC (Cooperative Patent Classification) code scheme, which was launched by the EPO and US patent office in 2012, with the aim to extend the IPC. The differences between IPC and CPC are only minor, and both schemes are updated regularly to account for emerging and changing technologies. The responsibility to determine the appropriate codes for a patent lies with the patent examiner.

A brief primer on quantum information technologies: The interdisciplinary research field of QI consists of a large number of subfields with varying potential for practical applications and degree of experimental realization. Notable subfields are quantum computation, quantum cryptography, quantum teleportation and quantum sensors & metrology. Quantum computation holds the potential to drastically speed up certain types of computations, such as prime number factorization. Although fully operational quantum computers are still some time away, simple proof of concepts do exist, and the controversial D-Wave Two prototype is allegedly capable of solving very specific problems by means of quantum phenomena. Quantum cryptography is perhaps the most mature field, has strong economic potential (e.g. in the financial industry), and commercial products are already available from startups such as MagiQ and ID Quantique. Quantum teleportation, on the other hand, is rather exotic. In itself it is unlikely to hold any economic potential for the foreseeable future, mostly because of the insurmountable challenge to experimentally realize quantum state teleportation for more than just a few atoms. It could however be used as an auxiliary technique, e.g. for entanglement swapping. Finally, quantum sensors & metrology are a more pragmatic field, with the aim to use quantum entanglement for the measurement of variables, such as magnetic fields. As such, this field is interesting for practical applications in unrelated fields, such as medicine.

Of course, neither quantum algorithms (Shor, Grover, ...) and protocols (BB84, E91, SARG04, ...) nor the general idea of quantum computing itself (one-way, adiabatic, topological, ...) can be patented. What can be - and is - patented is the concrete implementation. In particular, patent applicants have to decide which physical media to work with (laser pulses, optical lattices, trapped ions, NMR, ...).

 

Secondary Sources

Some literature on patent activity in the area of QI already exists: The UK patent office, IPO (Intellectual Property Office), has released two very informative reports on quantum technologies in 2013 and 2014, respectively (available here and here). Furthermore, in 2013 Winiarczyk et al. published a paper "Analysis of patent activity in the field of quantum information processing" in the peer-reviewed IJQI journal (preprint available here).

The 2013 IPO report breaks down its analysis into the following three categories: quantum secure communications, quantum computation, quantum metrology & sensors. A fourth category, quantum simulators, was excluded from detailed analysis due to lack of patents. The 2014 IPO report analyses four slightly different categories: quantum telecommunications, quantum computation, quantum sensors, quantum timing & atomic clocks. Although the fourth category clearly qualifies as a quantum technology, it has only a small overlap with QI, which is why I will disregard this category in the following.

The IJQI paper analyses 523 patents from 1993 to 2011. This paper is insightful, but unfortunately seems to contain a mistake. Since the authors were not aware that patents become publicly available only 18 months after filing date, they found a sharp decrease of new patents for the years 2010 and 2011, the two years preceding their 2012 study. No such drop in patent activity is reported in the two IPO reports. I've e-mailed the authors, and they told me that they have recently come to the same conclusion, and they will consider doing another study.

 

IPC and CPC codes

One way to find QI patents is to search the IPC and CPC databases for relevant keywords such as 'quantum cryptography' in the title, abstract or first page of a patent description. Alternatively, one could browse through the relevant IPC or CPC codes. The latter strategy is complicated by the fact that QI-related patents are dispersed over a wide range of disparate categories. This can be seen in the following diagram, where the most frequent codes of QI-related patents are displayed.

quantum information patents

(download PDF)

Clearly visible is the hierarchical structuring of the IPC and CPC into sections, classes, subclasses, groups, and subgroups. The smallest items, the subgroups, may branch out into other subgroups. Groups or subgroups that were only recently added are marked by their date of addition between brackets at the end of their description. Groups and subgroups popular with patents relating to one of the three QI subfields 'quantum computing', 'quantum sensors & metrology', and 'quantum communication' (which includes quantum cryptography and quantum key distribution) are denoted by colors. The more intense the color, the more commonly the group or subgroup is used. Color combinations arise when more than one QI subfield is relevant.

The diagram also takes the differences between IPC and CPC into account: Where the CPC extends the description of an item, the extended description is displayed within curly brackets. Subgroups that only exist in the CPC scheme are framed by light gray rectangles.

Keywords that are associated with QI are displayed in bold font. It can be seen that there exist no classes or subclasses dedicated to QI technologies. Only on the group and subgroup level do keywords appear, and in some cases only in the CPC scheme. From this background it is not surprising that the authors of the IJQI paper find that the IPC codes of QI-related patents are dispersed over a wide range of subclasses: they categorize around 20% of codes as belonging to 'other' subclasses, which is around the same percentage as that of the codes belonging to the single largest subclass, namely H04L. Because of this, only the most common codes could be included in the above diagram. Some less common codes not displayed here are group G02B and the subgroups H03K19/195 and G06E3/00.

Something interesting can be seen at groups B82Y10 and B82Y20. Although their descriptions include QI-related keywords, only a negligible amount of relevant patents can be found in these two groups. Nevertheless, they were included in the diagram, because of their relevant description. On a final note, many of the 'quantum sensors & metrology' patents are only loosely related to QI processing. This especially true for class G01, whereas the patents in group H01L39 are more tangibly related to QI in the narrower sense.

 

Timeline breakdown

Throughout the 1990s the amount of newly granted QI-related patents stagnated on a very low level. From the year 2000 onwards there has been a sharp increase that reached its peak in around 2005. It should be mentioned that two of the biggest contestants, D-Wave Systems and MagiQ, were both founded in 1999. Since 2005, the amount of new patents has remained relatively stable, with trends visible only in subfields:

  • Quantum Communication: peaked in 2007 at around 270 patents; has since evened out to around 200 new patents annually
  • Quantum Computing: peaked in 2005 at 85 patents; has since slightly declined to current level of around 60 new patents annually
  • Quantum Sensors & Metrology: uninterrupted upwards trend, albeit from low level; currently around 20 new patents annually

 

Patent holder breakdown

The major QI patent holders can be partitioned into three categories, of whom the first two categories are by far the most dominant:

  • dedicated QI startups: D-Wave Systems, MagiQ, ...
  • established IT corporations: Hewlett-Packard, Toshiba, Microsoft, IBM, NEC, NTT, ...
  • academic institutions: University of Michigan, John Hopkins University, Chinese University of Hong Kong, ...

With regard to the subfields, the holders of the most patents are:

  • Quantum Communication: MagiQ, NEC, Toshiba, Hewlett-Packard, NTT, Mitsubishi
  • Quantum Computing: D-Wave Systems, Hewlett-Packard, Toshiba, Japan Science & Technology Agency, NTT, Microsoft
  • Quantum Sensors & Metrology: Hitachi, Sumitomo, Toshiba, University of Michigan

Notably absent here are ID Quantique and Google, two companies with a strong record of involvement in QI technologies. ID Quantique is a Switzerland-based quantum cryptography startup and direct competitor of MagiQ. Google has its own QI research team (the Google Quantum A.I. Lab) and recently acquired a D-Wave Two prototype in cooperation with NASA and others.

 

Country breakdown

The dominant countries from where patent applications are filed are:

  • Quantum Communication: Japan, USA, Europe
  • Quantum Computing: USA, Canada, Europe, Japan
  • Quantum Sensors & Metrology: USA, Europe, Japan

Nothing surprising here. The strong position of Canada in quantum computing is no doubt due to D-Wave Systems.

The countries in which a patent comes into force is determined by the patent authority with which the application is filed. For QI-related patents the most popular authorities are, in that order, the US patent office, the WIPO, the EPO, and the Japan patent office. Interestingly, with the exception of the UK patent office, the national patent offices of Europe receive only a negligible amount of QI-related applications. Apparently, applicants prefer to take the route via the EPO.

 

Disclaimer: While every care has been taken in the compilation of the information contained in this article, I cannot, and do not, guarantee its accuracy, completeness, usefulness, up-to-dateness or availability. In particular, the contents of this article do not constitute legal advice, and you cannot rely on the information as an alternative to legal advice from an attorney or similar legal services provider. I will not be held responsible for any claim, loss, damage or inconvenience caused as a result of any information or download within this article.