In the global race to build bigger and better quantum computers, researchers have taken a step forward. A new machine called Helios is a radically different system compared to other quantum computers.

Quantum computers harness the power of quantum mechanics, the laws that govern how physics operates at atomic and sub-atomic scales. Among various designs for such machines, Helios is a trapped-ion quantum computer, which means it uses charged atoms suspended in free space using electromagnetic fields.

It operates using 98 qubits – the units of information that a quantum computer uses to process data. This number of qubits makes it the largest trapped-ion quantum computer built so far. Quantinuum, the company behind the device, which is based in Cambridge, UK and Broomfield, Colorado, demonstrated earlier machines operating on 32 qubits in 2023 and 56 qubits in 2025.

Helios and its predecessors use an architecture (or operational structure) with separate regions for storing and processing quantum information. The architecture is called a QCCD (quantum charge-coupled device) and was invented in 2002. This is akin to the architecture of classical computers that have a memory for storage (hard disk drives, solid state drives) and a separate processor (CPU, GPU).

The specific geometry of Helios resembles a rosette, with the ring for storage and two streamers for processing. The crucial element is the four-way X junction where they meet. A QCCD physically transports the charged objects (the ions) from the storage to the processing regions electrically.

The processing is done using laser pulses. This requires a quantum algorithm to be broken up into ion transport and quantum processing. This differs substantially from other quantum computing architectures such as superconducting qubits, where the qubits are fixed in space, and processing is executed by electrical signals switched on and off in time.

Error correction

Classical charge-coupled devices, the template for the types now used in Helios, were the central component in early digital cameras. There are three main advantages to now using these in quantum computation.

Gates are the fundamental building blocks of quantum circuits, which operate on qubits. One advantage of the QCCD approach is that gates operating on certain qubits do not affect their neighbours. This reduces a phenomenon known as “crosstalk”, leading to better quality performance.

The second advantage is the ability to measure and reset qubits during a computation. This enables inevitable errors in the computation to be detected and corrected at the earliest possible time, again improving performance.

Finally, as qubits can be accessed from and returned to very different parts of the memory after processing (within limitations), distant qubits can be connected, leading to more efficient operation.

The Helios device demonstrated all the above advantages. It relies on two major advances – one in hardware and one in software. First, the four-way X junction lets the system handle several tasks at once rather than one at a time, which is much faster. That was impossible in earlier QCCD machines, which could only move data back and forth in a single line or loop.

Secondly, judicious use of the freedom offered by the two dimensions of the X junction relies on a new classical control software called Helios runtime which plans the smartest, fastest route for moving and processing the data.

Together, these represent a substantial advance in the engineering of quantum computers. An important consequence is that Helios can perform computations that cannot be performed even on the largest supercomputers using known methods within reasonable amounts of time and power consumption.

Computational prowess

It shows the ability of this device, and its successors, in surpassing the computational prowess of classical computers – albeit their algorithms and hardware are rapidly advancing too.

Having said that, Helios’ computations have just been random benchmark tests, so the practical importance of this leap forward is still limited. To perform quantum computations of practical importance from areas of science and commerce, even the most optimistic estimates suggest that quantum computers must be enormously bigger and better – of the order of a million qubits.

The UK National Quantum Strategy Mission 1 proposes to have a fault-tolerant quantum computer (FTQC) with a trillion operations. For reference, the longest computation Helios ran had about 4,000 operations operating on 98 qubits.

Building FTQCs out of the Helios device, or any other known architecture or system for that matter, remains a formidable challenge. As the numbers suggest, it would require connecting thousands of QCCD devices, each with thousands of qubits, via quantum links.

Visions exist for QCCD devices with a two-dimensional grid layout resembling city streets that can operate on thousands or more qubits. The inevitable challenge of engineering and operating such a system with an immense number of junctions and concomitant qubit-jams – akin to traffic jams – is yet to be encountered, let alone overcome. The physical transporting of qubits is also a very slow process because it uses very precise voltages, and this will only grow with larger devices and more qubits.

As of now, the Helios device is an exciting advance in the field of quantum computing. Whatever form the FTQCs of the future take, they will involve imaginatively engineered quantum hardware devices assisted by smart software.

Animesh Datta receives funding from the UKRI, Horizon Europe programme, UK Space Agency, and European Space Agency.

Global air travel is predicted to more than double by 2050, threatening to exceed the capacity of air traffic control officers (atcos) to oversee our skies safely.

One solution would be to recruit more controllers – but that’s not as easy as it sounds. Airspace in the UK and elsewhere is broken up into geographically and altitude-bounded sectors – with each sector normally the responsibility of one atco (the tactical controller) at any time. So, increasing the number of controllers to deal with a greater concentration of aircraft could require a major redesign of UK airspace.

An alternative is to move beyond the current generation of air traffic control technology by increasing the role of AI systems – including providing real-time decision support for atcos.

AI agents could continuously monitor each airspace sector and recommend instructions (known as “clearances”) to issue to aircraft in that sector. These would then be assessed for their appropriateness by the tactical controller.

The UK’s air traffic control provider NATS (a public-private partnership) is, like other national providers, on an automation journey. It has collaborated with the University of Exeter and the Alan Turing Institute on Project Bluebird, which is investigating how AI systems could help meet the challenge of growing air travel volumes.

---

AI has long been discussed as a threat to jobs and livelihoods. But what’s the reality? In this series, we explore the impact it is already having on different occupations – and how people really feel about their AI assistants.

---

As part of this project, I have been using immersive participant observation and interviews to investigate how atcos use their existing tools, explore how they feel about the adoption of AI, and how the trust relationships that are key to safe air traffic operation could be affected.

As one controller reflected about existing air traffic procedures: “It’s an overall improved system when you have more trust in the people that you’re working with, and you’re more familiar with them.”

Trust in the AI

Atcos are organised into “watches” – groups of about 15 controllers who regularly work together on the same shift in the same control centre to oversee sectors within their region. While each sector is the responsibility of one tactical controller, the movement of aircraft between sectors requires seamless teamwork.

Automated short-term conflict alert systems to prevent air collisions have been a key part of air traffic control for several decades now. This offers an example of how AI tools may be received, as one atco recalled:

We were all quite nervous when that [conflict alert] system came out – we didn’t know how it was going to work. But very soon, you get your confidence in the system. You understand that the predictions that have been given to you are correct 98% of the time, or whatever they are – and you learn the small minority of situations where it isn’t.

From my discussions with atcos, it is clear that if AI agents are to be trusted, their recommendations must be presented in ways that controllers can quickly make sense of, and decide if they are appropriate. One told me:

There could be all sorts of possibilities where an AI agent is becoming your support controller, your secondary controller, and making suggestions you can either accept or decline … it’s a very interesting concept, if it can demonstrate to me why that’s a good idea – because that’s the bit we don’t know.

Concerns were raised with me about AI agents being “black boxes” whose workings lack transparency. This could make it difficult for an atco to determine when AI recommendations should be accepted and – just as important – when they should not be.

Delivering transparency means satisfying two requirements: explainability (knowing how the AI system reached its decision) and interpretability (understanding why the AI system reached that decision). As part of the project, we have identified ways to ensure transparency – such as the AI agent always revealing which aircraft within a sector it is taking into account when recommending clearances, and why.

In short, teaming atcos with AI agents could play an important role in handling growing air traffic volumes safely – but only if atcos develop strong levels of trust in the AI agents first.

A digital twin

Project Bluebird is working on more efficiency benefits for air traffic control and other air traffic management services. These include more accurate methods for aircraft trajectory prediction, using models trained on tens of thousands of observations of how aircraft behave in flight.

It is building a digital twin of UK airspace – a virtual model of airspace sectors, routes, procedures and other factors that form the operational context for air traffic. This high-fidelity model could be used for the training and evaluation of new types of AI agent, and testing the viability of new airspace designs.

Bluebird is also developing an assurance framework that regulators and other air traffic control stakeholders could use to determine if the safety requirements for AI-based systems are met.

These systems are likely to play an increasing role if the projected growth in air traffic is to be managed safely. While NATS is committed to keeping humans in the loop for the foreseeable future, the atcos I spoke to are cautiously optimistic about the benefits of AI. As one observed:

[Unlike a tactical controller,] an AI agent can think of the whole network. It can look at the whole trajectory of a flight … and even a small adjustment to an aircraft speed or heading can make a big difference to the overall flight.

Rob Procter receives funding from UK Research and Innovation.

Humans are not the only animals that communicate. Birds sing to attract mates, dolphins use signature whistles that identify individuals, and many primates produce complex vocalisations that help them coordinate their social lives.

By studying these communication systems, researchers can reconstruct some of the evolutionary steps that eventually led to one of our defining traits: language.

Language doesn’t just mean making sounds – it requires an extraordinary ability to control them. Human speech demands precise coordination between breathing, the vocal folds across the larynx, and our tongue, lips and jaw.

Humans’ closest living relatives – chimpanzees and bonobos – possess vocal tracts that are more similar to ours than scientists once believed. Yet they do not seem to have the same fine voluntary control over the structures – starting with our lips and tongue – that allow us to produce rapid sequences of speech and such diverse sounds.

This raises an intriguing question: when did our ancient ancestors begin to acquire this remarkable vocal control? One possible clue comes from an unexpected source: laughter.

The evolution of laughter

For several years, I have studied the rhythmic structure of vocalisations in non-human primates. Rather than focusing on what calls mean, I have been interested in how they are produced.

The timing of vocalisations can reveal how breathing and vocal movements are coordinated, offering a window into the mechanisms that generate them. Whether analysing the songs of Madagascar’s indris lemur or the calls of orangutans, I kept coming back to the same question: how much control do different primates have over the rhythm of their voices?

Laughter turned out to be an ideal system for exploring this question.

Unlike speech, our ability to laugh is shared with all the great apes. Chimpanzees, bonobos, gorillas and orangutans all produce laughter-like vocalisations during play. Although these sound quite different from our familiar “ha ha ha”, they serve similar social functions, signalling playful intentions and strengthening social bonds.

Human laughter is much more diverse, however. We can produce loud belly laughs, quiet chuckles, giggles, snorts or almost silent laughter. We laugh when we are amused but also when we are embarrassed, nervous or trying to ease social tension. This flexibility makes laughter much more than a simple emotional reflex for us.

Human laughter is highly context-dependent. We laugh differently when we are genuinely amused, when we engage in polite social interactions, when we feel awkward, or when we share an inside joke.

Together with colleagues, I analysed 140 laughter bouts from humans and the four other great ape species. Instead of asking whether laughter sounded similar across species, we examined its rhythm: how quickly each laugh element followed the previous one, and how consistent that timing was. Our results revealed a clear evolutionary pattern.

When laughing, humans show greater temporal variability than our closest relatives. Chimpanzees and bonobos produce laughter at a slower rhythm than humans, while gorillas and orangutans laugh at an even slower pace.

In contrast, human laughter consists of more rapidly repeated elements – successive vocal bursts that together form a laugh. This increased flexibility may be one of the evolutionary changes associated with the emergence of finer voluntary control over vocal production.

In other words, humans’ ability to “laugh fast” may reflect one of the key innovations that eventually made speech possible.

The link between laughter and speech

The evolution of speech may not simply have involved producing new sounds, but changing the very tempo at which vocal actions are organised. Our findings suggest traces of this key evolutionary transition are still preserved in one of our most familiar activities: laughing.

While humans and chimpanzees share a common evolutionary history, we will never hear how our ancient ancestors laughed. Fossils preserve bones not vocalisations, so the only way to reconstruct the evolution of laughter is by comparing living species and inferring the changes that occurred along the hominin lineage.

Because we found that chimpanzee and bonobo laughter differs from human laughter in rhythmic structure and temporal variability, we can use these living species to infer that early hominins may have exhibited laughter patterns which were already distinct from those of other great apes during the evolutionary transition towards modern human laughter.

Three-to-four million years ago, early hominins may already have shown some of the foundations of the flexible laughter patterns seen in humans today. Our findings suggest that increased temporal variability in laughter evolved along the hominin lineage. This may have enabled laughter to become increasingly adapted to different social contexts, from playful interactions to maintaining social bonds and signalling affiliation.

Every time we compare ourselves with our closest great ape relatives, we discover that the gap between humans and other apes is smaller than was once imagined. Yet we also uncover subtle differences that may have had profound evolutionary consequences.

Laughter is one of them. It may seem like a simple expression of joy, but hidden within its rhythm is a record of millions of years of evolution – and, perhaps, one of the clues to how our human ancestors eventually found their voice.

Chiara De Gregorio does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

Millions of people will take to the roads this holiday season, only to end up spending frustrating hours sitting in traffic jams. Congestion costs drivers time, fuel and patience – while also increasing pollution and placing huge pressure on transport networks.

If you’ve ever found yourself staring enviously at the lane next to you, convinced it’s moving faster, you’re not alone. Most of us instinctively believe that changing lanes will get us home sooner. But mathematics suggests this intuition is usually wrong.

As an applied mathematician, much of my research is driven by one question: how can we predict and control the behaviour of complex systems operating under uncertainty? And this can apply as much to holiday traffic as it does to emerging quantum technologies.

We tend to think traffic jams need a trigger: an accident, roadworks or a lane closure. Yet some of the most frustrating queues have no obvious cause at all. These are known as phantom traffic jams – stop-start waves that travel backwards through traffic even though every vehicle is moving forwards.

In 2008, Japanese researchers asked 22 drivers to drive around a circular track while maintaining a constant speed and safe distance from the vehicle in front. Within minutes, tiny and unavoidable differences in braking and reaction times grew into a stop-start wave that travelled continuously around the circuit, despite there being no obstacle anywhere on the road.

The explanation is surprisingly simple. One driver brakes slightly more than necessary. The driver behind reacts a little later and brakes a little harder. The next driver does the same.

In this way, the initial tiny disturbance grows as it travels backwards until drivers hundreds of metres behind are forced to stop completely, even though nobody can identify what caused the queue in the first place.

The mathematics behind traffic

Rather than modelling every driver individually, mathematicians treat traffic as a continuous flow – borrowing ideas from fluid dynamics, where the movement of vehicles is analysed much like the flow of water through a pipe.

One of the simplest relationships in traffic flow theory is q = ρv, where q is the traffic flow (the number of vehicles passing a point each hour), ρ is the traffic density (amount of cars on the road) and v is their average speed.

This deceptively simple equation explains a counterintuitive phenomenon. Initially, adding more vehicles obviously increases the overall traffic flow, as more vehicles pass along the road.

But once the road becomes too crowded, everyone is forced to slow down. Eventually, this reduction in speed outweighs the increase in number of vehicles such that the overall flow is reduced.

The equation shows there is an optimal traffic density that maximises the number of vehicles passing through the road each hour. Beyond that point, adding more cars reduces the efficiency of the road – and increases the time it takes everyone to get to their destination.

The same mathematics explains why constantly changing lanes is rarely worthwhile.

A lane change creates a small disturbance that neighbouring drivers must react to. If many drivers behave in the same way, these disturbances accumulate and increase the likelihood of traffic waves. So, what feels like a clever decision for one driver can ultimately make conditions worse for everyone.

Can maths help reduce traffic jams?

In my research on probabilistic mathematical methods, I develop approaches that combine prediction, coordinated decision-making and feedback to keep complex systems stable, even when the available information is incomplete or “noisy” (full of extraneous data).

In traffic, the disturbances we seek to prevent are the stop-start waves that create phantom traffic jams. In other complex systems, they might be power failures, communication bottlenecks or unstable autonomous systems.

Rather than designing solutions for one specific application, applied mathematicians develop general mathematical frameworks. That’s one reason why ideas originally developed for engineering systems can also help us think differently about traffic – with each of our mathematical tools playing a different role:

1. Signal processing transforms measurements from road sensors, cameras and connected vehicles into useful information.

2. Stochastic modelling accounts for uncertainty arising from driver behaviour, weather conditions and changing traffic demand.

3. Machine learning identifies patterns in all this data, and forecasts where congestion is likely to develop.

4. Control theory then closes the loop by determining how traffic systems should respond, whether by adjusting traffic-light timings, introducing adaptive speed limits, recommending alternative routes or coordinating fleets of autonomous vehicles.

Rather than optimising the journey for one driver, the objective becomes improving the performance of the entire traffic network.

Perhaps the most striking demonstration came in 2018, when researchers repeated the Japanese circular-track experiment with one key difference. They replaced one of the human-driven vehicles with a single autonomous car, programmed to accelerate and brake smoothly.

Remarkably, that one vehicle, less than 5% of the traffic, was enough to dampen the stop-start wave around the entire circuit, improving traffic flow and reducing fuel consumption for every driver involved.

Tips for your next car journey

Traffic congestion is unlikely ever to disappear completely. Population growth, increasing travel demands and the unpredictability of human behaviour will always put pressure on our roads.

But mathematics is helping us move away from reacting to congestion, towards preventing it.

One important aspect is understanding how large networks of interacting agents can make coordinated decisions using only local information, without requiring every component to know the state of the entire system. These ideas are central to the design of intelligent transport systems, where connected vehicles and infrastructure must cooperate to improve traffic flow.

In the meantime, if you’re about to set off on holiday in your car, here are three driving tips all backed up by strong mathematical evidence. If everyone adopts them, this should reduce the time you spend in traffic queues.

• Maintain a safe following distance;

• Accelerate and brake smoothly;

• Resist the temptation to keep switching lanes in search of tiny gains.

Applied mathematics shows the fastest way to your destination isn’t to drive more aggressively. It’s to help keep the entire system stable.

Randa Herzallah have received external research funding from the Leverhulme Trust ("Control of Behavioural Dynamics"), the Engineering and Physical Sciences Research Council (EPSRC), and Isansys (ECG-derived biosignals PhD studentship, 2016–2019)