Robert F. Kennedy Jr. says he’ll make big changes in the US government if Donald Trump is elected. He may find them hard to pull off, no matter what position he’s appointed to.
Physicists and others with STEM backgrounds are sought after in industry for their analytical skills. However, traditional training in STEM subjects is often lacking when it comes to nurturing the soft skills that are needed to succeed in managerial and leadership positions.
Our guest in this podcast is Peter Hirst, who is Senior Associate Dean, Executive Education at the MIT Sloan School of Management. He explains how MIT Sloan works with executives to ensure that they efficiently and effectively acquire the skills and knowledge needed to be effective leaders.
This podcast is sponsored by the MIT Sloan School of Management
New field experiments carried out by physicists in California’s Sierra Nevada mountains suggest that intermittent bursts of embers play an unexpectedly large role in the spread of wildfires, calling into question some aspects of previous fire models. While this is not the first study to highlight the importance of embers, it does indicate that standard modelling tools used to predict wildfire spread may need to be modified to account for these rare but high-impact events.
Embers form during a wildfire due to a combination of heat, wind and flames. Once lofted into the air, they can travel long distances and may trigger new “spot fires” when they land. Understanding ember behaviour is therefore important for predicting how a wildfire will spread and helping emergency services limit infrastructure damage and prevent loss of life.
Watching it burn
In their field experiments, Tirtha Banerjee and colleagues at the University of California Irvine built a “pile fire” – essentially a bonfire fuelled by a representative mixture of needles, branches, pinecones and pieces of wood from ponderosa pine and Douglas fir trees – in the foothills of the Sierra Nevada mountains. A high-frequency (120 frames per second) camera recorded the fire’s behaviour for 20 minutes, and the researchers placed aluminium baking trays around it to collect the embers it ejected.
After they extinguished the pile fire, the researchers brought the ember samples back to the laboratory and measured their size, shape and density. Footage from the camera enabled them to estimate the fire’s intensity based on its height. They also used a technique called particle tracking velocimetry to follow firebrands and calculate their trajectories, velocities and accelerations.
Highly intermittent ember generation
Based on the footage, the team concluded that ember generation is highly intermittent, with occasional bursts containing orders of magnitude more embers than were ejected at baseline. Existing models do not capture such behaviour well, says Alec Petersen, an experimental fluid dynamicist at UC Irvine and lead author of a Physics of Fluids paper on the experiment. In particular, he explains that models with a low computational cost often make simplifications in characterizing embers, especially with regards to fire plumes and ember shapes. This means that while they can predict how far an average firebrand with a certain size and shape will travel, the accuracy of those predictions is poor.
“Although we care about the average behaviour, we also want to know more about outliers,” he says. “It only takes a single ember to ignite a spot fire.”
As an example of such an outlier, Petersen notes that sometimes a strong updraft from a fire plume coincides with the fire emitting a large number of embers. Similar phenomena occur in many types of turbulent flows, including atmospheric winds as well as buoyant fire plumes, and they are characterized by statistically infrequent but extreme fluctuations in velocity. While these fluctuations are rare, they could partially explain why the team observed large (>1mm) firebrands travelling further than models predict, he tells Physics World.
This is important, Petersen adds, because large embers are precisely the ones with enough thermal energy to start spot fires. “Given enough chances, even statistically unlikely events can become probable, and we need to take such events into account,” he says.
New models, fresh measurements
The researchers now hope to reformulate operational models to do just this, but they acknowledge that this will be challenging. “Predicting spot fire risk is difficult and we’re only just scratching the surface of what needs to be included for accurate and useful predictions that can help first responders,” Petersen says.
They also plan to do more experiments in conjunction with a consortium of fire researchers that Banerjee set up. Beginning in November, when temperatures in California are cooler and the wildfire risk is lower, members of the new iFirenet consortium plan to collaborate on a large-scale field campaign at the UC Berkeley Research Forests. “We’ll have tonnes of research groups out there, measuring all sorts of parameters for our various projects,” Petersen says. “We’ll be trying to refine our firebrand tracking experiments too, using multiple cameras to track them in 3D, hopefully supplemented with a thermal camera to measure their temperatures.
“My background is in measuring and describing the complex dynamics of particles carried by turbulent flows,” Petersen continues. “I don’t have the same deep expertise studying fires that I do in experimental fluid dynamics, so it’s always a challenge to learn the best practices of a new field and to familiarize yourself with the great research folks have done in the past and are doing now. But that’s what makes studying fluid dynamics so satisfying – it touches so many corners of our society and world, there’s always something new to learn.”
The particle physics community is in the vanguard of a global effort to realize the potential of quantum computing hardware and software for all manner of hitherto intractable research problems across the natural sciences. The end-game? A paradigm shift – dubbed “quantum advantage” – where calculations that are unattainable or extremely expensive on classical machines become possible, and practical, with quantum computers.
A case study in this regard is the Institute of High Energy Physics (IHEP), the largest basic science laboratory in China and part of the Chinese Academy of Sciences. Headquartered in Beijing, IHEP hosts a multidisciplinary scientific programme spanning elementary particle physics, astrophysics as well as the planning, design and construction of large-scale accelerator projects – among them the China Spallation Neutron Source, which launched in 2018, and the High Energy Photon Source, due to come online in 2025.
Quantum opportunity
Notwithstanding its ongoing investment in experimental infrastructure, IHEP is increasingly turning its attention to the application of quantum computing and quantum machine-learning technologies to accelerate research discovery. In short, exploring use-cases in theoretical and experimental particle physics where quantum approaches promise game-changing scientific breakthroughs. A core partner in this endeavour is Shandong University (SDU) Institute of Frontier and Interdisciplinary Science, home to another of China’s top-tier research programmes in high-energy physics (HEP).
With senior backing from Weidong Li and Xingtao Huang – physics professors at IHEP and SDU, respectively – the two laboratories began collaborating on the applications of quantum science and technology in summer 2022. This was followed by the establishment of a joint working group 12 months later. Operationally, the Quantum Computing for Simulation and Reconstruction (QC4SimRec) initiative comprises eight faculty members (drawn from both institutes) and is supported by a multidisciplinary team of two postdoctoral scientists and five PhD students.
Hideki Okawa “IHEP and SDU are creating a global player in the application of quantum computing and quantum machine-learning to HEP problems.” (Courtesy: IHEP)
“QC4SimRec is part of IHEP’s at-scale quantum computing effort, tapping into cutting-edge resource and capability from a network of academic and industry partners across China,” explains Hideki Okawa, a professor who heads up quantum applications research at IHEP (as well as co-chairing QC4SimRec alongside Teng Li, an associate professor in SDU’s Institute of Frontier and Interdisciplinary Science). “The partnership with SDU is a logical progression,” he adds, “building on a track-record of successful collaboration between the two centres in areas like high-performance computing, offline software and machine-learning applications for a variety of HEP experiments.”
Right now, Okawa, Teng Li and the QC4SimRec team are set on expanding the scope of their joint research activity. One principal line of enquiry focuses on detector simulation – i.e. simulating the particle shower development in the calorimeter, which is one of the most demanding tasks for the central processing unit (CPU) in collider experiments. Other early-stage applications include particle tracking, particle identification, and analysis of the fundamental physics of particle dynamics and collision.
“Working together in QC4SimRec,” explains Okawa, “IHEP and SDU are intent on creating a global player in the application of quantum computing and quantum machine-learning to HEP problems.”
Sustained scientific impact, of course, is contingent on recruiting the brightest and best talent in quantum hardware and software, with IHEP’s near-term focus directed towards engaging early-career scientists, whether from domestic or international institutions. “IHEP is very supportive in this regard,” adds Okawa, “and provides free Chinese language courses to fast-track the integration of international scientists. It also helps that our bi-weekly QC4SimRec working group meetings are held in English.”
A high-energy partnership
Around 700 km south-east of Beijing, the QC4SimRec research effort at SDU is overseen by Xingtao Huang, dean of the university’s Institute of Frontier and Interdisciplinary Science and an internationally recognized expert in machine-learning technologies and offline software for data processing and analysis in particle physics.
“There’s huge potential upside for quantum technologies in HEP,” he explains. In the next few years, for example, QC4SimRec will apply innovative quantum approaches to build on SDU’s pre-existing interdisciplinary collaborations with IHEP across a range of HEP initiatives – including the Beijing Spectrometer III (BESIII), the Jiangmen Underground Neutrino Observatory (JUNO) and the Circular Electron-Positron Collider (CEPC).
Big science, quantum advantage QC4SimRec will apply innovative quantum approaches to build on SDU’s pre-existing interdisciplinary collaborations with IHEP across a range of HEP initiatives, including the Jiangmen Underground Neutrino Observatory (above). (Courtesy: Yuexiang Liu)
One early-stage QC4SimRec project evaluated quantum machine-learning techniques for the identification and discrimination of muon and pion particles within the BESIII detector. Comparison with traditional machine-learning approaches shows equivalent performance on the same datasets and, by extension, the feasibility of applying quantum machine-learning to data analysis in next-generation collider experiments.
“This is a significant result,” explains Huang, “not least because particle identification – the identification of charged-particle species in the detector – is one of the biggest challenges in HEP experiments.”
Xingtao Huang “Our long-term goal is to establish a joint national laboratory with dedicated quantum computing facilities across both campuses.” (Courtesy: SDU)
Huang is currently seeking to recruit senior-level scientists with quantum and HEP expertise from Europe and North America, building on a well-established faculty team of 48 staff members (32 of them full professors) working on HEP. “We have several open faculty positions at SDU in quantum computing and quantum machine-learning,” he notes. “We’re also interested in recruiting talented postdoctoral researchers with quantum know-how.”
As a signal of intent, and to raise awareness of SDU’s global ambitions in quantum science and technology, Huang and colleagues hosted a three-day workshop (co-chaired by IHEP) last summer to promote the applications of quantum computing and classical/quantum machine-learning in particle physics. With over 100 attendees and speakers attending the inaugural event, including several prominent international participants, a successful follow-on workshop was held in Changchun earlier this year, with planning well under way for the next instalment in 2025.
Along a related coordinate, SDU has launched a series of online tutorials to support aspiring Masters and PhD students keen to further their studies in the applications of quantum computing and quantum machine-learning within HEP.
“Quantum computing is a hot topic, but there’s still a relatively small community of scientists and engineers working on HEP applications,” concludes Huang. “Working together, IHEP and SDU are building the interdisciplinary capacity in quantum science and technology to accelerate frontier research in particle physics. Our long-term goal is to establish a joint national laboratory with dedicated quantum computing facilities across both campuses.”
One thing is clear: the QC4SimRec collaboration offers ambitious quantum scientists a unique opportunity to progress alongside China’s burgeoning quantum ecosystem – an industry, moreover, that’s being heavily backed by sustained public and private investment. “For researchers who want to be at the cutting edge in quantum science and HEP, China is as good a place as any,” Okawa concludes.
For further information about QC4SimRec opportunities, please contact Hideki Okawa at IHEP or Xingtao Huang at SDU.
Quantum machine-learning for accelerated discovery
To understand the potential for quantum advantage in specific HEP contexts, QC4SimRec scientists are currently working on “rediscovering” the exotic particle Zc(3900) using quantum machine-learning techniques.
In terms of the back-story: Zc(3900) is an exotic subatomic particle made up of quarks (the building blocks of protons and neutrons) and believed to be the first tetraquark state observed experimentally – an observation that, in the process, deepened our understanding of quantum chromodynamics (QCD). The particle was discovered in 2013 using the BESIII detector at the Beijing Electron-Positron Collider (BEPCII), with independent observation by the Belle experiment at Japan’s KEK particle physics laboratory.
As part of their study, the IHEP- SDU team deployed the so-called Quantum Support Vector Machine algorithm (a quantum variant of a classical algorithm) for the training along with simulated signals of Zc(3900) and randomly selected events from the real BESIII data as backgrounds.
Using the quantum machine-learning approach, the performance is competitive versus classical machine-learning systems – though, crucially, with a smaller training dataset and fewer data features. Investigations are ongoing to demonstrate enhanced signal sensitivity with quantum computing – work that could ultimately point the way to the discovery of new exotic particles in future experiments.
Optical traps and tweezers can be used to capture and manipulate particles using non-contact forces. A focused beam of light allows precise control over the position of and force applied to an object, at the micron scale or below, enabling particles to be pulled and captured by the beam.
Optical manipulation techniques are garnering increased interest for biological applications. Researchers from Massachusetts Institute of Technology (MIT) have now developed a miniature, chip-based optical trap that acts as a “tractor beam” for studying DNA, classifying cells and investigating disease mechanisms. The device – which is small enough to fit in your hand – is made from a silicon-photonics chip and can manipulate particles up to 5 mm away from the chip surface, while maintaining a sterile environment for cells.
The promise of integrated optical tweezers
Integrated optical trapping provides a compact route to accessible optical manipulation compared with bulk optical tweezers, and has already been demonstrated using planar waveguides, optical resonators and plasmonic devices. However, many such tweezers can only trap particles directly on (or within several microns of) the chip’s surface and only offer passive trapping.
To make optical traps sterile for cell research, 150-µm thick glass coverslips are required. However, the short focal heights of many integrated optical tweezers means that the light beams can’t penetrate into standard sample chambers. Because such devices can only trap particles a few microns above the chip, they are incompatible with biological research that requires particles and cells to be trapped at much larger distances from the chip’s surface.
With current approaches, the only way to overcome this is to remove the cells and place them on the surface of the chip itself. This process contaminates the chip, however, meaning that each chip must be discarded after use and a new chip used for every experiment.
Trapping device for biological particles
Lead author Tal Sneh and colleagues developed an integrated optical phased array (OPA) that can focus emitted light at a specific point in the radiative near field of the chip. To date, many OPA devices have been motivated by LiDAR and optical communications applications, so their capabilities were limited to steering light beams in the far field using linear phase gradients. However, this approach does not generate the tightly focused beam required for optical trapping.
In their new approach, the MIT researchers used semiconductor manufacturing processes to fabricate a series of micro-antennas onto the chip. By creating specific phase patterns for each antenna, the researchers found that they could generate a tightly focused beam of light.
Each antenna’s optical signal was also tightly controlled by varying the input laser wavelength to provide an active spatial tuning for tweezing particles. The focused light beam emitted by the chip could therefore be shaped and steered to capture particles located millimetres above the surface of the chip, making it suitable for biological studies.
The researchers used the OPA tweezers to optically steer and non-mechanically trap polystyrene microparticles at up to 5 mm above the chip’s surface. They also demonstrated stretching of mouse lymphoblast cells, in the first known cell experiment to use single-beam integrated optical tweezers.
The researchers point out that this is the first demonstration of trapping particles over millimetre ranges, with the operating distance of the new device orders of magnitude greater than other integrated optical tweezers. Plasmonic, waveguide and resonator tweezers, for example, can only operate at 1 µm above the surface, while microlens-based tweezers have been able to operate at 20 µm distances.
Importantly, the device is completely reusable and biocompatible, because the biological samples can be trapped and undergo manipulation while remaining within a sterile coverslip. This ensures that both the biological media and the chip stay free from contamination without needing complex microfluidics packaging.
The work in this study provides a new type of modality for integrated optical tweezers, expanding their use into the biological domain to perform experiments on proteins and DNA, for example, as well as to sort and manipulate cells.
The researchers say that they hope to build on this research by creating a device with an adjustable focal height for the light beam, as well as introduce multiple trap sites to manipulate biological particles in more complex ways and employ the device to examine more biological systems.
Connexion
