3 min read
Part of the success of quantum technology lies in finding ways to work with imperfect hardware. But that same imperfection gives quantum technology an uncomfortable relationship with absolute concepts such as “same” and “different”. And the flipside of successfully exploiting imperfect hardware can become a real challenge when it comes to patenting the resulting technology: does the invention rely on quantum systems having identical states or properties? What if, in physical reality, they sometimes differ?
For example, two qubits may be designed to behave identically yet, as a matter of practical reality, their realised physical properties are more likely to be slightly different. Consider two qubits intended to operate at the same frequency, but whose realised frequencies are 5.000 GHz and 5.001 GHz. Looking at the precise figures, those frequencies are evidently different. But does that difference represent a meaningful technical difference between the qubits? Or should they be regarded as effectively the same, taking into account the inevitable imperfections of the hardware?
When it comes to protecting the technology, this raises the question: how do we distinguish an accidental difference arising from physical imperfection from an intentional difference that forms part of the invention? A patent protecting quantum technology should have a good answer to that question. Recently, Rigetti's European patent EP3111381 (see here) was challenged on this point and, as it turns out, has a good answer to that question.
Rigetti's patent relates to a quantum system in which neighbouring pairs of qubits have “distinct qubit operating frequencies”. Following grant, the patent was challenged on the basis of a similar earlier quantum system. This earlier quantum system was designed to operate with identical frequencies but, it was argued, necessarily has “distinct qubit operating frequencies” because manufacturing tolerances are a matter of practical reality. And so the argument was that the qubit operating frequencies must be “unavoidably different” and that this amounts to “distinct qubit operating frequencies”.
But do “unavoidably different” qubit operating frequencies equate to “distinct” qubit operating frequencies? The European Patent Office ruled that those are not the same thing. The Office accepted that manufacturing gives rise to a spread in the frequencies of nominally identical qubits, but reasoned that the claimed “distinct” frequencies go beyond those normal manufacturing variations.
So where is the line between incidentally different and deliberately different? Rigetti's patent connects the distinct frequencies to the technical effect of reducing noise. This gives the claimed difference a technical context. The point is not simply that two numerical values happen to be unequal. The frequencies have been differentiated within an engineered system for a technical purpose. Understanding the invention in this context gives meaning to “distinct” that goes beyond frequencies that merely happen to be different.
Manufacturing tolerances are only one example of a much wider issue in quantum technology. Imprecision can arise through control and calibration. State preparation occurs only with finite fidelity. Environmental interactions, noise and decoherence create further departures between an intended system and its physical realisation. Measurement introduces yet another layer. And operating frequency is just one example. The underlying issue is much wider and can arises in relation to many physical properties of quantum systems. The specific challenge brought against Rigetti's patent is therefore representative of a much broader issue for quantum technology.
This raises the question whether using “distinct”, rather than “different”, provides a general solution. Put differently: is “distinct” a magic word that carries intrinsic meaning? And are there any other magic words to use when protecting quantum inventions? No. Even in Rigetti's patent, the word “distinct” was not treated in isolation. Instead, the Office considered the technical context in which the claimed invention operates.
The technical context can generally help to determine what words such as “different” or “distinct” mean for a particular invention. For example, the achieved technical effect can indicate that a specified difference is deliberate and technically meaningful, rather than merely an incidental consequence of imperfect hardware. But there are limits to how much work that context can do. In Rigetti's case, the claimed language aligned with the interpretation that the frequencies are not merely different but are intentionally differentiated in a way that is technically relevant to noise reduction. That worked because of careful use of claim language and precise description of the technology.
The challenge outlined here is in principle applicable to more than just quantum technology, but its relevance to quantum technology is particularly striking because quantum technology is limited by both practical engineering and fundamental physics. Finding ways to build useful systems from components that are noisy, variable and imperfectly controlled plays a critical part in the technological advances in the quantum sector. And a patent protecting such innovation needs to conceptually distinguish the deliberately engineered solution from the underlying, imperfect physical system into which it is embedded.
The appropriate approach will, of course, vary between inventions. As general guidance:
The desired technical effect can provide important context. It can help establish the underlying intention and inform how the language defining the invention should be understood.
Context is no replacement for precise patent language. Great care is still needed to describe the relevant technical distinction as accurately as possible.
In some cases, difficulties arising from practical implementation of the invention can be reduced by moving one level of abstraction upwards. Consider, for example, a control system configured to prepare two qubits in identical states. The intention is that both qubits are put into identical states when practical reality may be that they are not in fact in the same state. That can provide an elegant way of protecting some quantum inventions, but it is important to recognise that abstraction brings its own risks. In particular, the boundary between abstract quantum physics and its physical implementation needs careful consideration when it comes to patenting quantum inventions before the European Patent Office. I discussed that related issue in more detail here.
Working in our engineering team, Johannes primarily advises clients who devise technical inventions falling within the field of mechanical engineering or related fields. He prepares and prosecutes patent applications before patent offices in the UK and overseas, primarily Europe and the US. Johannes also advises his clients in relation to contentious issues, particularly with a view to settling disputes out of court. Additionally, Johannes has experience with proceedings before the UK Patent Office relating to matters of inventorship and (in-)validity.
Email: johannes.biniok@mewburn.com
Stay up to date with our latest thinking.