Content area
Purpose
The purpose of this paper is to provide details of developments in quantum technologies and consider their potential applications in robotics.
Design/methodology/approach
Following a short introduction, this study first provides an overview of the global quantum technology landscape. It then discusses developments in quantum computing and sensing technologies. Potential applications in robotics are then considered and finally, brief conclusions are drawn.
Findings
Quantum technologies are the topic of a rapidly growing global R&D effort. Quantum computing has the potential to conduct conventional computations far more rapidly than traditional computers and solve complex problems that are presently challenging or impossible. If realised, robotic applications could include enhanced route planning, machine learning and data fusion. Quantum position and magnetic field sensors have the potential to revolutionise navigation systems in airborne, land and marine robots and overcome limitations of GPS and inertial measurement units. Magnetic sensors also have a role in health care in the control of robotic prostheses and exoskeletons and in brain–computer interface techniques. Quantum radar, lidar and imaging systems stand to outperform their conventional counterparts, and applications are anticipated in military and civilian robots. Quantum technologies are still at an early stage of development, and much progress will be made in the future, opening up many further robotic applications.
Originality/value
This study provides an insight into quantum technology developments and their potential applications in robotics.
Introduction
Quantum theory is the branch of physics that seeks to explain phenomena occurring at the atomic or sub-atomic scale and the formative work was conducted by Bohr, Planck, Einstein and others during the early years of the 20th century. The two main quantum principles driving developments in quantum technology (QT) are entanglement and superposition. Briefly, quantum entanglement is when two atoms or particles are connected, or “entangled”, despite being spatially separated. If you change the properties of one, the other changes instantly. Quantum superposition is the idea that particles simultaneously exist in multiple states. When a measurement is performed, it is as if the particle selects one of the states in the superposition. These seemingly paradoxical concepts cannot be understood through classical physics.
Numerous devices have been developed that are underpinned by quantum phenomena, such as lasers, flash memories and LEDs. These are not regarded as QT devices as they do not involve the direct manipulation of quantum states. A quantum device can be defined as one that intentionally uses or harnesses quantum mechanical effects for its operation. These have the potential to yield devices and techniques that outperform their conventional counterparts and, perhaps, offer unique functionalities and examples include quantum computers, sensors, imaging devices and communications. Most are still at the research or early prototype stages and are the topic of global academic and corporate development efforts. The aim of this article is to discuss the state of these activities and provide a preliminary insight into their potential use in robotics.
Overview of the global quantum technology landscape
QT is viewed as being potentially disruptive or strategic, and all of the world’s leading industrialised nations are involved with research and developments. Table 1 provides an insight into the scale and nature of these activities and illustrates that the USA and China presently have the greatest involvement, albeit through greatly differing approaches. China has provided massive state funding to research, whereas the USA has seen very significant investment in QT start-ups. In addition, both Canada and the UK have seen strong levels of start-up investment, at around $1.3bn and $1.6bn, respectively (2001–2023) and the UK’s National Quantum Strategy, announced in 2023, has committed £2.5bn to developing and commercialising QT over the next 10 years.
The USA is leading the quantum computing field, and in addition to its world-class academic institutions, companies such as IBM, Google, Intel, Microsoft and Amazon are heavily involved with the technology. The Chinese Government has declared QT to be a strategic priority and plans to establish a national quantum laboratory and a $10bn quantum research centre and the country has a clear edge in quantum communication. The $15bn public investment is part of China’s 14th five-year plan (2021–2025) and may well be extended during the next plan. India is an emerging player and approved its National Mission on Quantum Technologies and Applications in 2023, which allocated about $800m for quantum R&D over eight years. South Korea filed over 2,000 QT patents between 2000 and 2023, and the government announced a public investment of $2.3bn in 2023.
In addition to established companies, approximately 350 QT start-ups have been founded globally since 2000 (Figure 1), with at least 70% operating in the quantum computing space. Between 2000 and 2013, the number founded each year was in single figures but rapidly increased from 2014, peaking at approximately 60 in 2016. Since then it has fallen but remains in double figures (13 in 2023). The probable reasons for this decline are that investors are favouring AI start-ups, in part reflecting AI’s demonstrable commercial potential, and many also prefer to invest in QT scale-ups and later-stage businesses rather than new ventures.
Over 180,000 QT-related papers have been published since 1990 by workers from around 150 countries but approximately 75% originated from just 10: the USA, China, Canada, the UK, Germany, France, Italy, Japan, India and Russia. As shown in Table 1, China and the EU were the most prolific in 2023. Publications on quantum computing dominate and constitute around 70% of the total. Research and development activities range from underlying theory through software and component development, fabrication techniques, signal processing and novel materials to the production and evaluation of prototypes. The majority of developments concern three key applications: computing, sensing and communications.
Quantum computing
The concept of using quantum phenomena in computing arose in the 1980s, and Figure 2 illustrates the fundamental differences between conventional and quantum computers. The bit used in conventional computing is replaced by the qubit (quantum bit), and while bits can represent either 1 or 0, qubits can simultaneously represent 1, 0 or any value in between. This property allows a quantum computer to perform many operations simultaneously and in parallel, theoretically allowing computations to be performed many orders of magnitude faster than, or which are impossible with, classical computers. Qubits can be based on superconductivity, photons, neutral atoms, trapped ions and other structures, and most require cooling to a few mK to eliminate thermal noise and vibrations, which tend to destroy the information. Accordingly, most quantum computers feature complex cooling systems, as illustrated in Figure 3. The errors in classical bits can be corrected by copying the information to multiple bits, but because a quantum state cannot be copied, multiple physical qubits are entangled to form a logical qubit to reduce the error. As yet it is unclear which, if any, existing qubit classes will ultimately prevail.
The present-day state of operational systems is illustrated by IBM’s developments. In 2019, the company launched the Quantum System One, which was the first circuit-based commercial quantum computer. This is housed in an airtight borosilicate glass cube, each face being 2.7 m wide and tall, that maintains a controlled environment. A dilution refrigerator contains a 20-qubit superconducting transmon (transmission line–shunted plasma oscillation qubit) quantum processor. Figure 4 shows a Quantum System One at the Cleveland Clinic, a multispecialty academic medical centre. It was installed in 2023 and is the world’s first quantum computer to be dedicated to healthcare research. In late 2023, IBM launched the Quantum System Two, which is the first modular quantum computer system and contains three IBM Quantum Heron processors. The Heron is a 133-qubit, tuneable-coupler quantum processor that IBM claims eliminates crosstalk errors that emerged in its previous processors. IBM plans to introduce a system with 200 logical qubits capable of running 100 million quantum gates by 2029, with a goal of two thousand logical qubits running one billion gates by 2033.
Despite the extensive global R&D effort by major IT companies and academic institutions during recent decades, quantum computers remain at an early stage of development and are yet to demonstrate unambiguously their superiority over their conventional counterparts in all but a small number of highly specialised cases.
Quantum sensors
Sensors based on various quantum phenomena have been demonstrated that respond to a range of physical variables and have the potential to measure them with high levels of accuracy. The majority are still at an early stage of development.
Cold atom interferometry combines principles of quantum mechanics with laser cooling and interferometry. Atoms, typically of rubidium, are laser cooled to temperatures of just a few µK and then optically or magnetically trapped. They are then exposed to a sequence of laser pulses that act like beam splitters in an optical interferometer and create a superposition state, effectively splitting each atom into two different paths simultaneously. After splitting, the atoms travel along two different paths, during which time they acquire a phase difference based on the external forces they encounter. Another set of laser pulses is used to recombine the atoms, which causes interference between the two paths, and the interference pattern depends on the phase difference accumulated during their journey. Analysis of this pattern allows the extremely accurate measurement of the external forces.
Systems can be configured to measure multi-axis accelerations and rotations, thereby acting as quantum accelerometers and gyroscopes. Because of their accuracy, they have the potential to form the basis of advanced navigation systems by overcoming drift and other problems associated with the sensors used in conventional inertial measurement units (IMUs) or the limitations of GPS, such as jamming and connectivity loss. Most systems are large, complex and costly and feature numerous discrete optical and optoelectronic components, and to achieve more widespread uses, a level of miniaturisation and component integration is required. An example is work at Sandia National Laboratories, which has come some way to achieving this by fabricating the integrated optical structure shown schematically in Figure 5.
When nitrogen atoms replace carbon atoms in a diamond lattice, they form nitrogen-vacancy (NV) centres which have quantum spin states that are highly sensitive to magnetic fields. When green laser light is shone on the diamond, it excites the electrons within the NV centres to a higher energy level, and as they return to their ground state, they emit red light photons. Microwave radiation modulates this effect, and by varying the frequency of the radiation and observing changes in the intensity of the emitted red light, the external magnetic field intensity can be precisely determined. Sensors using this technique can resolve magnetic fields in the low nT to pT regions. Figure 6 shows a prototype quantum magnetic sensor developed by Bosch Quantum Sensing, a recently founded Bosch subsidiary. As with cold atom interferometry, academic groups are aiming to achieve high levels of miniaturisation and component integration.
Quantum imaging is an emerging family of techniques that exploits phenomena such as quantum entanglement to image objects with a resolution or other features that are beyond the capabilities of classical optics. Some examples are quantum ghost imaging, imaging with undetected photons and sub-shot noise imaging. Ghost imaging captures images of objects using photons that never interacted with the object itself and relies on the spatial correlations between pairs of photons generated through spontaneous parametric down-conversion (SPDC) within a non-linear medium. Figure 7 shows schematics of two ghost imaging configurations.
Quantum radar is an emerging technique that would operate by generating billions of entangled photon pairs, sending one photon from each pair into the search area while retaining the other in a memory (Figure 8). “Signal” photons reflected back to the sensor would then be compared to their “idler” counterparts, revealing information about the target. There are significant engineering challenges to be overcome, but unlike conventional radar, such sensors promise to be largely immune to jamming, and their use possibly undetectable. Accordingly, they have strategic importance and are under development by the US and Chinese military. Conceptually similar quantum lidar is also under development.
Applications in robotics
The perception that quantum computers will significantly outperform their traditional counterparts presently remains unproven. The following discussion assumes that this will ultimately be realised but must nevertheless remain speculative.
Quantum computers appear to have significant potential in solving optimisation problems. An example is determining the most effective path for an autonomous robot in a changing, unpredictable and/or cluttered environment. Quantum computing potentially solves these types of problems much better than classical computing, as quantum algorithms explore numerous solutions simultaneously. The Quantum Approximate Optimisation Algorithm which is designed to find approximate solutions to hard combinatorial optimisation problems on quantum computers can potentially find optimal or near-optimal solutions faster than classical algorithms, thereby allowing robots to plan routes more efficiently. Volkswagen has conducted trials with quantum routing algorithms and a D-Wave quantum computer to optimise traffic flow in cities. Another potential use is in machine learning in applications such as object recognition, decision-making and natural language processing. Various quantum machine learning algorithms have been developed and could speed up the training and inference phases and improve pattern recognition, enabling robots to learn and adapt in real time or process very large data sets far more rapidly than conventional systems. A quantum algorithm with potential in a number of robotic applications is Grover’s Algorithm. This addresses the problem of searching for an item in an unsorted database, which it achieves in a significantly reduced time compared to classical algorithms. It creates a superposition of all possible states of the database and then applies a series of quantum operations to amplify the amplitude of the state that contains the desired item and repeats this process until it finds the item with a high probability. Quantum computing may also find uses in data fusion in applications such as autonomous vehicles, where large quantities of data from multiple sensors need to be processed in real time.
The prospects for quantum sensors are more clear-cut, as the technologies are generally more advanced. Systems based on cold atom interferometry are poised to exert a major influence on navigation, as there is a real need for technology that is immune to jamming. Several companies are now commercialising this technology, and an example is M Squared, which has developed the UK’s first commercial 3D quantum accelerometer through a collaboration with Imperial College London. Another is US-based AOSense. The company’s six-axis quantum IMU was developed in conjunction with Boeing and was used in a test flight by Boeing in 2024, enabling the aircraft to navigate without GPS for 4 h. A system developed in the UK by Aquark Technologies, a spinout of the University of Southampton, has recently been used during navigation trials onboard HMS Pursuer by the Royal Navy (Figure 9) and in 2023 the system, which weighs less than 10 kg, was deployed on a quadcopter drone. The company’s patented laser cooling method of constructing cold atom quantum sensors without magnetic fields allows them to reduce the sensor’s size, weight, power consumption and cost. As the technology matures and applications become widespread, it is anticipated that cold atom interferometry systems will be deployed on mobile robots operating in the air, on land and at sea.
Quantum magnetic field sensors based on NV centres offer prospects to aid navigation as the Earth’s crustal magnetic field exhibits immutable and geographically unique patterns. Bosch is collaborating with Airbus to assess this possibility with its technology, which may ultimately supplement or replace GPS. However, it may face competition from cold atom interferometry, but like this technology, it has potential for mobile robotic application in all three media. There are also applications in robotic health care. These sensors will be able to detect the magnetic signals associated with muscle contractions, both from the upper and the deeper muscle layers, which are difficult to detect with myoelectric sensors. As such, they have prospects for the improved control of robotic prostheses and exoskeletons. Furthermore, by sensing the small magnetic fields associated with the brain’s electrical activity, these sensors could form the basis of magnetic brain–computer interface (BCI) systems, which, like their electrical counterparts, could allow individuals with paralysis to control robotic prostheses.
Quantum radar could theoretically detect targets up to four times faster than traditional radar and also sense small, slow-moving objects and operate in cluttered environments. Accordingly, applications are likely in military robots, both for target acquisition and threat detection. Equally, these capabilities make them well suited to uses in driverless cars and other autonomous land robots. This is equally true for quantum lidar, which will exhibit better temporal and spatial resolution than conventional lidar. The prospects for the various quantum imaging techniques are less clear, but the potential to create high-resolution images of objects obscured by dust, clouds or smoke again suggests applications in military robots and perhaps also in self-driving cars.
Conclusions
Quantum technologies are the topic of a rapidly growing global R&D effort, with the USA and China being the most active nations. Quantum computing is attracting the greatest interest and has the potential to conduct conventional computations far more rapidly than traditional computers and also solve complex problems that are currently challenging or impossible. While these capabilities presently remain unproven, if realised robotic applications could include enhanced mobile robot route planning, machine learning and data fusion. Sensors based on quantum phenomena are at a more advanced stage. Position sensors based on cold atom interferometry and magnetic field sensors exploiting NV centre diamonds have the potential to revolutionise navigation systems in airborne, land and marine robots in military and civil applications and to overcome the limitations of GPS and IMUs. Quantum diamond sensors also have a role in health care in the control of robotic prostheses and exoskeletons and in BCI techniques. Quantum radar, lidar and imaging systems stand to outperform their conventional counterparts, and applications are anticipated in military and commercial robots.
This article had provided a preliminary view of the prospects for QT in robots. The technologies are still at an early stage of development, and it is inevitable that much progress will be made in the future, opening up many further robotic applications.
Figure 1Total global QT start-ups, 2001–2023
Figure 2Comparison between conventional and quantum computers
Figure 3A dilution quantum refrigerator
Figure 4The IBM quantum system one at Cleveland Clinic
Figure 5Conceptual image of a fully integrated, multi-channel silicon photonic single-sideband modulator chip generating tones applied to a cold atom interferometer experiment powered by a single integrated laser source
Figure 6The NV centre-based quantum magnetic sensor
Figure 7Ghost imaging systems can be configured to take advantage of either the position (a) or momentum (b) correlations inherent in SPDC
Figure 8Schematic of quantum radar
Figure 9The quantum sensor onboard HMS pursuer during sea trials
Table 1
Overview of national QT activities
| USA | China | EU | Japan | |
|---|---|---|---|---|
| Proportion of global QT patent applications1 (%) | ∼25 | ∼25 | ∼20 | ∼23 |
| Proportion of global QT publications (2023, %) | 17.4 | 23.0 | 24.3 | 4.6 |
| Total public funding to 2022 ($bn) | 3.7 | 15.3 | 8.4 | 1.8 |
| Total investment in start-up companies2 ($m) | 3300 | 304 | 499 | ∼60 |
Notes:12020–2023; 22001–2022
© Emerald Publishing Limited.