Challenges in the application of systems engineering to quantum technologies

The nature of quantum technology is broad and ranges from quantum random number generators based on quantum states of light to quantum gravity sensors based on clouds of laser-cooled atoms. The nature of these individual systems themselves will affect the nature of the challenges posed to systems engineering.

Quantum enhanced communications

Quantum-enhanced communications systems include applications where a quantum component has been substituted for a classical one, with the quantum effects being contained within the quantum component itself; the effects on the wider system are more akin to existing technologies.

For example, in the case where a Rydberg atom ^{[note 1]} based receiver is used as an alternative to a classical antenna, quantum effects would be confined within the Rydberg subsystem. In this case, quantum effects are localised and need only be considered within a small subset of the overall system architecture.

Quantum enabled communications

In contrast to a quantum-enhanced system, a quantum-enabled system is fundamentally based on quantum principles. An example of such a system would be a future quantum information network based on entanglement and teleportation effects. In this case, the entire system is underpinned by quantum effects. From a systems engineering standpoint, these effects will pervade the entire system leading to more far-reaching consequences.

It is expected that out of these two cases, the quantum-enabled systems will pose a greater challenge to existing systems engineering methods. Some of the quantum effects that pose a challenge to conventional systems engineering techniques are summarised in Table 1 below.

More information on the quantum effects listed in this table can be found in references ^[E] and ^[F].

Table 1: Examples of quantum effects that will challenge systems engineering methodology. This table is not intended to represent a comprehensive list but rather to provide some examples of such effects

Some possible means of addressing these challenges include:

• Compartmentalisation – designing the architecture of the system in a sufficiently modular way
• Isolation – designing facilities into the system so that quantum sub-systems or components can be verified independently of the other parts of the system and their characteristics are self-contained.

Another challenge which affects the systems engineering of new and novel technologies, including quantum technologies, is the lack of suitable specialised models.

Until these are developed, the systems engineering of quantum systems will be generally more challenging. For example, within quantum systems, system execution cannot be recorded at run-time. This could be a challenge for assurance since traditional methods to audit the system’s performance may not be possible.         

Systems engineering and the exploitation of novel technologies

Systems engineering is an extensive technical field that not only sets out established practice but also provides guidance on how particular domains should be addressed and particular technologies included. This guidance includes the introduction of novel and disruptive technologies and the considerations that need to be addressed with these, where special measures may need to be adopted in order to ensure that they are successfully integrated into the wider system ^[12].  Whilst software is an increasingly dominant technology ‘substrate’ in many systems, other substrates may be incorporated and exploited in systems using systems engineering.

Such systems engineering guidance could result in a number of strategies being devised to incorporate and exploit quantum technologies. These strategies should seek to isolate or contain quantum-specific considerations from standard systems considerations at the macroscopic scale, should they be sufficiently different. This methodology would be particularly appropriate for quantum-enhanced systems where the quantum component is limited.

For example, a quantum sensor may be treated as a delicate instrument with appropriate interfacing and protection. Such an approach ensures that it can operate effectively and that any quantum effects are contained within that component in the system. Adopting this approach should allow the use of established systems engineering technologies to be continued as far as possible, with new techniques being required only for the quantum aspects of the systems and their interfaces.

Isolation and containment form a fundamental overall strategy for hybridising technologies including those that are quantum-based with those that aren’t – and the engineering associated with those technologies and the systems engineering which ‘spans’ them. This has yet to be widely done for systems that include quantum technologies. Therefore, the challenge here will be in applying this methodology to quantum systems that have a large number of distinct features that need to be taken into consideration.

For example, one of these distinct features is the minute scale of quantum phenomena. Quantum effects typically occur on scales on the order of 10^-8 m to 10^-16 m. Therefore, models used for macro-scale engineering will need to be made compatible with models of quantum phenomena.

This is not a unique problem and is an active area of research in other fields such as systems engineering of nanotechnology ^[13], but it has yet to be fully resolved.

Implications for requirements capture

Requirements capture is the result of engagement with users and other stakeholders in order to codify what the system needs to do and also what it needs to be. Requirements are important not only for recording the desired properties of the system, but also providing a means by which the final system can be contracted for, verified, and validated.

Requirements sets usually comprise both functional and non-functional requirements, which specify what the system must do and what it must be, respectively. Non-functional requirements typically relate to properties such as safety, security and reliability, for example.

Stakeholder requirements should be as far as possible solution-agnostic, however in reality the technology which will be used to provide a solution to the end users’ needs will often influence the form of requirements that should be stated. Quantum technologies will be no exception here.

For example, the most significant change could be a recognition that quantum technologies may exist that may enable the system to meet user requirements that have been excluded previously. As such using exploratory prototyping in development may help refine requirements for more novel quantum systems. The lack of widely used predictive models for quantum systems may pose a challenge in identifying the suitability of quantum solutions to a given requirements set. The effects of quantum technology on system requirements, however, will be much more impactful and require careful consideration. This will affect both functional requirements (due to potentially altered causation) and non-functional requirements (due to different characteristics).

Implications for system boundary analysis

For quantum-enhanced systems, system boundary analysis will be more akin to existing systems engineering methods than for quantum-enabled systems.

For example, in a quantum-enabled system where entanglement is present, the identification of the system boundary will become considerably more challenging. In particular, entanglement has the potential for information to flow around the system outside defined interfaces which will make a reductionist or “divide and conquer” approach less effective. Consequently, new strategies will need to be devised to accommodate quantum technologies. 

New strategies may entail working with multiple system boundaries, for example:

• one in which quantum effects dominate,
• one in which quantum effects are effectively contained,
• and a further zone in which some but not all quantum effects will need to be taken into account and will necessarily influence the style and scope of (systems) engineering that needs to be undertaken.

Implications for assurance and verification and validation

Verification and validation (V&V) is concerned with ensuring that the system that has been engineered satisfies the original requirements and stakeholder needs. For more sophisticated systems, especially those that are considered to be safety critical, this may lead to system certification, which will generally involve ensuring that the system is compliant with appropriate standards by organisations such as BSI or ISO/IEC, or defence standards, etc.

The V&V of quantum systems will be challenging as a result of the following four properties of quantum systems:

1. The observer effect: As with all systems, there is a need to design for tests; however, in the case of quantum technologies, certain tests will be impossible to carry out in practice. Due to the observer effect, measuring a quantum property can destroy it and render the measurement pointless. To address this problem, measurements will need to be conducted differently. For example, the system can be measured before being put into a quantum state and then after the quantum process has taken place to check consistency. Consequently, it is harder to build verification at all levels of the system in question.
2. Lack of determinism: At a fundamental level, the elements of a quantum system will not behave in a predictable manner, as is the case with their classical counterparts; research in systems engineering is considering the stochastic behaviour of systems due to the introduction of AI, and this may have applicability to quantum systems.
3. The presence of entanglement: Which leads to non-local phenomena and an ambiguous system boundary; will be more commonplace in quantum-enabled systems
4. State space explosion: The exponential growth of the dimensionality of the state space of a quantum system will mean the use of V&V techniques based on exhaustive searches are unlikely to be feasible. This could include exhaustive testing approaches as well as methods based on applying logic such as formal methods. While this is a problem for quantum technologies, it is not unique to them. Artificial intelligence-based technologies pose a similar challenge to system modelling, analysis and verification. In particular, the system’s capacity to learn exponentially increases the state space of the system. This effect also poses serious challenges to the verification of such systems, especially those that are to be used for high-integrity applications where safety and security are vital.

The nature and severity of the challenges posed by these four properties will also differ depending on the granularity of the V&V testing activity that is being considered. It is likely to be the case for some systems (e.g. optically pumped magnetometers) that quantum effects are more observable at the unit or component testing level as opposed to on a larger scale when the system has been integrated.

The presence of quantum effects will also likely lead to the presence of new failure modes that will need to be both identified and have mitigating strategies devised and adopted. This will include analysis at the quantum component level and combining the findings from such consideration with a similar analysis at the system level.

There are trends in several engineering areas (particularly in security engineering) to rely more strongly on ongoing evidence rather than ‘done and dusted’ approaches to V&V because this contributes to improved agility and responsiveness. Unless suitably robust methods for the assurance of quantum technologies can be developed, this could become a sticking point in the adoption of such technologies. One route to achieving this could be through carrying out DevOps with quantum system developers working closely with system integrators and test and evaluation experts.

Implications for reliability engineering

The purpose of reliability engineering is to ensure that the resulting system has the required reliability characteristics. The reliability of a system is inherently non-deterministic and is therefore analysed using statistical/probabilistic methods.

Quantum systems are likely to pose additional challenges for reliability engineering due to the potentially large number of failure modes, the use of entanglement, the decoherence of quantum states and the potentially non-deterministic nature of the components that comprise the system. It is, therefore, likely that new forms of analysis will be required to understand such properties and engineer them for reliability.

The containment of quantum effects: maximising the use of conventional systems engineering methodology

If conventional systems engineering methodologies are to be retained, then approaches and mechanisms for incorporating quantum technologies need to be devised and codified.

Systems engineering of quantum-based systems may need to be developed as a specific discipline in its own right; this would allow quantum technologies to be integrated into wider systems while allowing conventional systems engineering to be retained as far as possible.

The challenges are, therefore, twofold: (i) how to successfully engineer the quantum aspects of the systems and (ii) how to interface these with the remainder of the system consisting of more conventional technology substrates such as IT. 

This principle is shown conceptually in Figure 3 for quantum-enhanced and quantum-enabled communication systems  

A mapping is described below in Table 2 that shows the impact of quantum technologies on different stages of the systems engineering lifecycle according to the terminology that has been defined above.

Table 2: Quantum influence on lifecycle processes. The nature of the quantum system will determine how much of an impact there is on each lifecycle process, with some quantum systems likely to have very little impact, while others much greater on the specific process in question.  For example quantum clocks are expected to pose less of an integration challenge than quantum computers into existing systems.

References