Over the past months, a sudden influx of ‘Pokémon Go’ players could be observed across the globe. Youngsters, people of all ages scrutinise their surroundings silently, using their smartphones to catch those digital creatures with unlikely names. How could such a phenomenon take over the world so fast? Part of the answer may be the game’s strong interaction with the real-world and its impressive mapping, offering a whole new virtual experience of spaces that seem to be familiar and yet so different.Details
Based at the University of Warwick (UK), Sybille Lammes’ ERC-funded research focuses on digital mapping practices - including those related to play. She sees them as new media cultures that influence and alter our way of being and moving through spatial environments. She is also the first to combine New Media Studies, Science and Technology Studies and Human Geography and thereby to tackle an original and mostly unexplored research field.
“Digital maps can be simultaneously understood as new media, technologies and cartographies” explains Lammes, nourishing the conception of new media as “material cultures” that are physically embedded in daily life and technologies.
A new culture
Digital cartographies have changed the century-old conception of maps, traditionally perceived as something tangible representing a space. In the numerical world, on the other hand, we constantly interact with navigation and geolocation gadgets that are built in most applications and devices.
The maps of today are versatile real-time maps. Connected to online technologies – such as traffic sensors and GPS – they keep transforming before our eyes. Current navigation systems propose new routes when we miss a turn and adapt their suggestions to the way we play with them. We look for restaurants and shops through digital maps and some even date with the help of geolocation apps. Such daily tools influence our behaviour and thinking, the way we sense space and relate to it:
“The ambition of “Charting the Digital” is to define what digital mapping is about and what it culturally entails, and to explore, as a new techno-cultural phenomenon, its impact on our spatial relations”, says Lammes.
With her team, she has undertaken an extensive critical comparison of digital mapping practices, also in relation to traditional cartography and other related media forms. The research also encompasses the analysis of mapping interfaces, which act as technological mediators and induce users’ specific behaviour.
Sybille Lammes: “Our own research team members have become part of the study, by “going native”, i.e. becoming engaged and interacting with the material they study. They keep a journal about their experiences with the mapping interface and how they become acquainted with it”.
Digital mapping for playful learning
On the commercial side, Lammes’ team noted that highly successful and profitable digital mapping applications make use of a combination of gamification, connectivity and interactivity in their design. On this basis, they have developed a prototype of a location-based game: a mobile app named ‘Playfields’ that can be used to teach fieldwork to university students in a ‘hands on’ entertaining way.Project details:Researcher (PI):Sybille LammesHost institution:The University Of Warwick, United KingdomProject:Charting the Digital: Digital Mapping Practices as New Media Cultures, (Charting the Digital)ERC call:Starting Grant , ERC-2011-StG, panel SH5Max ERC funding:1,422,453 €Duration:Start date: 2011-11-01, End date: 2016-10-31
Despite recent advances in the fight against cancer, scientific research continues on several fronts. Current studies in the field of nanomedicine are proving very promising. Professor Valentina Cauda, from the Politecnico di Torino, has received funding from the European Research Council (ERC) for a pioneering project in this field, designed to develop therapies to target cancer cells without affecting the surrounding tissue.
Researcher picture: © Politecnico di TorinoDetails
What is your project about?
In my lab we design nanoparticles – known as TrojaNanoHorses (TNH) – which are engineered to be biomimetic, meaning that they can be injected into the bloodstream without causing immune reactions within the receiving body, that is, the tumour. The idea is that of the Greek myth of the Trojan Horse, which is not recognized as dangerous, hence is taken inside the city. Just like the horse, once inside the tumour cell, the nanoparticles give rise to mechanisms that can destroy it. In addition to this therapeutic effect, we have also designed the nanoparticles as diagnostic tools. In fact, when stimulated with ultraviolet lighting techniques, they will allow a better vision of the area affected by the tumour.
Currently there are many studies on this type of targeted treatment. What is the specificity of the nanoparticles you work with?
There are numerous and very valid studies carried out in recent years in the field of nanomaterials, materials of the dimensions of one billionth of a meter. Many of these involve the use of nanodrugs that are incorporated in the particle to be transported to the tumour cell. The difference of this project is that we do not foresee the use of chemotherapeutical substances associated with the nanoparticle. This should reduce the risks linked to the administration of these substances, in particular their unintentional release during the journey, which can damage other healthy tissues and organs. The therapeutic effect is due to a mechanism by which the nanoparticle disintegrates inside the tumour cell, releasing ions and radicals that damage it. Often, one hears of free radicals that are created by sunlight exposure and cause the ageing of cells. The procedure is very similar but, in this case, the release of radicals is extremely localized, and only occurs within the cancer cell.
Could this also be a valuable treatment for breast cancer?
The key to reaching the tumour is given by the molecules placed on the surface of the nanoparticle, which give the exact "address" to hit the target tumour cells. It will therefore be necessary to find the specific molecules to target breast cancer, in cooperation with clinicians and biologists with experience treating this type of tumour. It is essential, in fact, to find the peptide or antibody that binds to a specific receptor on the cancer cell, the selectivity otherwise fails and you can end hitting other cells. In general, it will be possible to extend these considerations to any type of cancer.
How will your project develop?
For the moment, the project is focused on the development of the nanoparticle in the lab. In the future, it will also cover the mechanisms of treatment and diagnosis on the tumour. During the next five years we will do studies on cultures of cancer cells in the lab and, if the data will be comforting, we can move on to clinical trials on patients, probably not before the next ten years.
How did you decide to specialise in nanomedicine?
Nanotechnologies have opened new horizons in various fields, as demonstrated by the recent Nobel Prize for chemistry about molecular machines. After my PhD, I started working with various groups that were working on nanoparticles for controlled drug delivery. Nanomedicine is an exciting interdisciplinary field, and without materials science, chemistry and physics, therapies may not be developed in an innovative and effective way. In Italy we are rather advanced in this field. In my group, eleven specialists from different disciplines work with me. They are all Italian, but we are trying to open up internationally thanks to an ERC initiative that will allow us to host researchers from around the world visiting an ERC laboratory.
Valentina Cauda is a chemical engineer. After obtaining her PhD in materials sciences in Italy, she worked at the University of Munich collaborating with various groups specialising in biomedical research. She decided to return to Turin initially at the Italian Institute of Technology and now, with ERC funding for her project, at the Politecnico di Torino, where she leads a research team.
Prof. Valentina Cauda and her team at Politecnico di Torino
Malgrado i recenti progressi nella lotta contro il cancro, la ricerca scientifica non si arresta e continua su vari fronti. Gli studi attualmente condotti nell’ambito della nanomedicina si stanno rivelando molto promettenti. La Prof.ssa Valentina Cauda, del Politecnico di Torino, ha ottenuto un finanziamento dallo European Research Council (ERC)per un progetto d’avanguardia in questo campo, mirato a sviluppare terapie che distruggono le cellule tumorali senza intaccare i tessuti circostanti.
In cosa consiste il suo progetto?
Nel mio laboratorio progettiamo delle nano-particelle - ribattezzate nano-Cavalli di Troia - che vengono ingegnerizzate per essere biomimetiche: questo significa che potenzialmente possono essere iniettate nel circolo sanguigno senza suscitare reazioni immunitarie nell'organismo che le riceve, cioè nel tumore. L'idea è quella del mito greco del Cavallo di Troia, che non essendo riconosciuto come pericoloso, viene lasciato entrare nella città. Proprio come il cavallo, una volta all’interno della cellula tumorale, le nanoparticelle danno luogo a meccanismi che la distruggono. Oltre a questo effetto terapeutico, le abbiamo progettate anche come strumenti diagnostici. Infatti le nanoparticelle, stimolate con tecniche di illuminazione ultravioletta, permetteranno anche di visualizzare l'area tumorale colpita.
Per saperne di più:
Attualmente esistono già molte altre ricerche su questo tipo di trattamento. Qual è la specificità delle nanoparticelle che studiate?
Ci sono tantissime e validissime ricerche condotte già da diversi anni nel campo dei nanomateriali, materiali di dimensioni nanometriche (1 miliardesimo di metro). Molte di queste prevedono l'utilizzo di nano-farmaci che vengono incorporati nella particella per essere trasportati fino alla cellula tumorale. La differenza di questo progetto è che noi non prevediamo l'utilizzo di chemioterapici associati alla nanoparticella. Questo dovrebbe abbattere i rischi legati alla somministrazione di tali sostanze, in particolare il rilascio involontario durante il percorso, che rischia di danneggiare altri tessuti e organi sani. L'effetto terapeutico è dovuto ad un meccanismo per cui la nanoparticella si disintegra all'interno della cellula tumorale liberando ioni e radicali che la danneggiano. Si sente spesso parlare dei radicali liberi che si creano quando ci si espone alla luce solare e che causano l'invecchiamento delle cellule. Il procedimento è molto simile ma, in questo caso, il rilascio di radicali è estremamente localizzato, e avviene soltanto all'interno della cellula tumorale.
Si tratta di un trattamento valido anche per i tumori al seno?
La chiave per raggiungere il tumore è data dalle molecole poste sulla superficie della nanoparticella e che danno “l'indirizzo” esatto per andare a colpire le cellule tumorali bersaglio. Sarà quindi necessario trovare “l’indirizzo” specifico per il cancro alla mammella, in collaborazione con medici e biologi esperti in questo tipo di tumore. È fondamentale infatti trovare il peptide o l'anticorpo che si lega al recettore presente sulla specifica cellula tumorale, altrimenti viene a mancare la selettività e si va a colpire qualsiasi tipo di cellula. In generale sarà possibile estendere queste considerazioni a qualsiasi tipo di tumore.
Per saperne di più:
Come si svilupperà il suo progetto?
Il progetto per ora riguarda lo sviluppo della nanoparticella in laboratorio, quindi lo studio dei meccanismi di terapia e diagnosi sul tumore. Durante i prossimi cinque anni faremo studi su culture di cellule tumorali in laboratorio e, se i dati saranno confortanti, si potrà passare alla sperimentazione clinica sui pazienti ma non prima di una decina anni.
Come mai ha deciso di specializzarsi in nanomedicina?
Le nanotecnologie aprono nuovi orizzonti in vari campi, come ha dimostrato anche il recente Premio Nobel per la Chimica sulle macchine molecolari. Dopo il dottorato ho cominciato a lavorare con vari gruppi specializzati su nanoparticelle per il rilascio controllato dei farmaci. La nanomedicina è un campo appassionante perché è interdisciplinare, e senza la scienza dei materiali, la chimica e la fisica, le terapie non potrebbero esser sviluppate in modo innovativo ed efficace. In Italia siamo abbastanza all’avanguardia in questo campo. Nel mio gruppo lavorano undici persone specializzate in ambiti diversi. Sono tutti italiani, ma stiamo cercando di aprirci a livello internazionale grazie ad un’iniziativa dell'ERC che permetterà di ospitare ricercatori da tutto il mondo in visita a un laboratorio ERC.
Per saperne di più:
Valentina Cauda è un ingegnere chimico. Dopo aver ottenuto il dottorato in scienza dei materiali in Italia, ha lavorato all’Università di Monaco di Baviera collaborando con vari gruppi attivi nella ricerca biomedica. Ha deciso di tornare a Torino prima presso l’Istituto Italiano di Tecnologia e ora, con il finanziamento del progetto ERC, presso il Politecnico di Torino dove dirige un’équipe di ricerca.Project details:Researcher (PI):Valentina CaudaHost institution:Politecnico Di Torino, ItalyProject:Hybrid immune-eluding nanocrystals as smart and active theranostic weapons against cancer, (TROJANANOHORSE)ERC call:Starting Grant , ERC-2015-STG, panelMax ERC funding:1,489,219 €Duration:Start date: 2016-03-01, End date: 2021-02-28
By focusing on certain actions and reactions within the brain, an EU-funded project has advanced understanding of how Alzheimer’s Disease develops. This could potentially open the door to a new era of targeted treatments.
Alzheimer’s Disease (AD) is the most common form of dementia, with symptoms that include gradually worsening memory loss and confusion. While there is no cure at present, a project funded by the European Research Council has opened up new research avenues that could one day lead to new therapeutic applications.
“Scientists have begun to better understand the genetics behind AD,” says MIRNA_AD project coordinator Bart De Strooper from VIB in Belgium. “For example, some 25 genes have been identified that we now know influence the genetic risk of developing the disease. We also know that amyloid peptides (a type of amino acid) are the main component of the amyloid plaques found in the brains of Alzheimer patients.”
Scientists believe that this plaque build-up blocks cell-to-cell signalling and could trigger inflammation. “What we don’t know, however, is exactly how Alzheimer’s progresses,” he adds.
Understanding disease pathways
De Strooper argues that the conventional focus on targeting amyloid peptides has been too simplistic. “It is based on the prediction that if you take away the amyloids, then you’ll have the cure,” he says. “The truth is that the development of AD is far more complex. It is not like killing the microbe that gives you pneumonia; AD involves a series of abnormal processes in the brain. What is needed is a new, non-linear way of thinking about how actions and reactions in the brain contribute towards the development of AD. We know now, for example, that abnormalities in amyloids start decades before the disease develops.”
The ERC grant enabled De Strooper to think outside of the box, and as a result make some important discoveries. “Our overall focus was on identifying pathways in the development of Alzheimer’s,” he explains. “We began by looking at both proteins and non-coding RNA (molecules involved in various biological roles in the expression of genes, but which are not translated into proteins), and examined the role that non-coding RNA might play in the development of AD.”
Potential therapeutic targets
The project was able to characterise important changes that take place in RNA in the brain of AD patients. In particular, De Strooper and his team identified several specific RNA molecules that changed significantly, such as the molecule micro-RNA 132. The next stage was then to investigate how these changes affect the brain in AD patients.
“By studying these micro-processes, we have arrived at a broader view of how AD develops, and achieved a broader understanding of what is going on in the brains of patients,” says De Strooper. “We now know that certain micro-RNAs regulate the amyloid and the Tau pathway together and could have an influence in inflammatory aspects of AD.”
These discoveries have opened up new research avenues. De Strooper is currently exploring whether micro-RNA molecules could be a potential target for therapeutic drugs, through studying the effect of manipulating these micro-RNA in mice. “It is perhaps a little early to say, but I think this research has the potential to be spun off one day and commercialised.”
The project has also underlined the complexity of Alzheimer’s, and the need to take a more holistic approach to research. “This is really where the ERC grant came into its own,” says De Strooper. “It can be difficult to get funding for this type of project, because they are seen as too speculative and risky. But the ERC is all about removing barriers and taking risks. The grant enabled me to carry out complex and efficient research, and I was able to train people who have since gone on to take up excellent positions.”Project details:Researcher (PI):Bart Geert Alfons Paul De StrooperHost institution:Vib, BelgiumProject:Role of microRNA dysregulation in Alzheimers Disease, (MIRNA_AD)ERC call:Advanced Grant , ERC-2010-AdG, panel LS5Max ERC funding:2,500,000 €Duration:Start date: 2011-05-01, End date: 2016-04-30
Many industries – and each of our cells – depend on emulsions. An EU-funded researcher has developed a method for studying molecules at the interface between oil nanodroplets and the water-based liquid contained in these substances. Her work advances understanding of liquid interfaces and emulsion stability, and is of great interest to industry.
Emulsions are mysterious phenomena – at the molecular level, at least. Found in industrial products such as lubricants, foods such as milk and in each of our cells, emulsions are made up of tiny droplets of one liquid that are dispersed in another.
The MINE project used an innovative spectroscopic light-scattering based method to study the molecules at the interface of these two liquids. This provides direct information about the molecules’ composition, orientation and environment – information that is useful for industry and developing new medicines.
“Our research can help us understand emulsions better and find better, cheaper ways to produce products,” says the project’s principal investigator, Sylvie Roke, head of the Laboratory for fundamental BioPhotonics at the École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland. She adds: “Companies from different industrial sectors are very interested in our work.”
Understanding the interface structure and the mechanics of its formation could boost drug development and understanding how drugs interact with the body’s cells. It could also reduce manufacturing's need for surfactants – substances that help oil and water blend in products such as perfumes.
Although scientists already know that the composition and structure of these interface molecules determines an emulsions’ properties (such as stability and reactivity), until now they had no way to study them directly. Emulsions are liquids that contain moving droplets, and existing molecular probes cannot reach or are not sensitive enough to focus on the one-molecule-thick interface layers around the droplets.
MINE’s method – developed by Roke – overcomes this barrier and has produced new information about how these molecules organise themselves.
The method combines non-linear optics and light scattering. Non-linear optics uses infrared laser light that lets molecules vibrate at the surface of a droplet. The vibrating electrons in the molecules allow the emission of light particles (known as photons) of a specific colour – but only when the molecules are at the droplet/liquid interface.
This light travels in a specific direction, measured by light scattering. It contains unique information about the composition of the interface – which molecules are present, how they are organised and whether they interact with water.
By analysing this information, Roke was able to update scientific views on liquids and especially emulsions. One fundamental question concerns charge: droplets or air bubbles in water move towards a positive electrode. This means they possess a negative charge. The common explanation is that hydroxide ions in the liquid are the source of this charge, yet Roke found this is incorrect, finding no such ions on droplet surfaces.
She also found that charged surfactants – that protect the fatty oil droplets from the water – form very different structures than traditionally expected. Rather than forming a film of densely packed molecules around the droplets, they form dilute films with only a surfactant molecule here and there.
These results are unexpected, says Roke. “But since we are not assuming anything, and these are direct measurements, they open up a new understanding of soft nanoscopic systems, which include emulsions. This will also be useful for other branches of nanotechnology, for example it will help to tailor food products that also have biological functions, for our health.”
MINE was fully funded with a starting grant from the European Research Council. “The funding was fantastic,” says Roke. “Building this technology was high risk/high gain and took a long time to get right. The ERC grant gave me time to answer fundamental questions.”
She and her team are now helping other labs to implement her techniques and models. She is also adapting her technique to study water droplets that are billionths of a metre across. “This will teach us more about how water behaves in small spaces, which is relevant for understanding the molecular architecture of our own bodies”Project details:Researcher (PI):Sylvie RokeHost institution:Ecole Polytechnique Federale De Lausanne, SwitzerlandProject:Molecular Interfacial structure and dynamics of Nanoscopic droplets in Emulsions (MINE), (MINE)ERC call:Starting Grant , ERC-2009-StG, panel PE4Max ERC funding:1,150,000 €Duration:Start date: 2009-11-01, End date: 2014-10-31
While on court, beach volleyball players need to act as a whole in order to prevent the ball from touching the sand: in a fraction of a second - just before the opponent's hand spikes the ball - the passer has to predict and adjust to the attacker's action as well as to their teammate's block position. Thanks to her Consolidator Grant, cognitive science professor Natalie Sebanz is studying the cognitive and psychological mechanisms underlying joint action expertise – in other words, how individuals learn skilled actions, such as those performed by professional athletes, together.Details
Many human achievements, from planning and executing architecture plans to performing surgeries, piano duets and tangoes, are the result of collaboration. According to Prof. Sebanz, teamwork is key for the advancement of human civilisation and heavily relies on joint actions. These occur whenever two or more people interact with one another to coordinate a particular action in space and time, in order to accomplish a shared goal.
Although philosophers agree that joint actions require shared intentions, debate on what shared intentions are is still ongoing. Besides, intending to do something together is clearly not enough to meet a target: "Imagine if a football team were to spend hours talking about how they will score a goal.There would be no guarantee that they would eventually manage to do so. This is why cognitive psychology and cognitive neuroscience are necessary to understand the mechanisms that come into play when people act together and allow for the fine-grained, timely coordination we see in team sports", explains Prof. Sebanz.
According to her research, members of a team rely on a variety of mechanisms to coordinate their actions. One crucial aspect is that, rather than focusing on their own specific movements, the team members mainly rely on the interaction with each other. Joint action expertise also involves planned coordination: "Take synchronous swimmers – their planned actions have to include their own contribution as well as their partners'. They all have an image of what their moves will look like when performed together. Being driven by this kind of mental imagery helps them to be so coordinated."
Prof. Sebanz's research also demonstrates that participation in highly coordinated activities boosts the participants’ sense of commitment: "The more athletes depend on each other, the more they feel bound to go on doing their part even though they are exhausted, like members of a rowing team who are running out of energy". This concept might also be the reason why rugby teams are so tightly knit – players value the group's performance and welfare more than their individual contribution.
In the project, Prof. Sebanz’s team also explores the benefits of joint improvisation. "Practising with exercises used in improvisational theatre, where participants cannot easily guess their partners’ intentions, may help team players to become more attuned to each other, while being confronted with an activity that is outside their field of expertise", concludes Prof. Sebanz.
The project, which started in 2014, uses electroencephalography along with behavioural and physiological indicators, such as measures of movement trajectories and heart rate synchronization. In the long term,
it may help develop autonomous robots designed to collaborate with humans and provide new therapies, based on social training interventions, to alleviate social disorders such as autism.Project details:Researcher (PI):Natalie SebanzHost institution:Kozep-Europai Egyetem, HungaryProject:Joint action expertise: Behavioral, cognitive, and neural mechanisms for joint action learning, (JAXPERTISE)ERC call:Consolidator Grants , ERC-2013-CoG, panel SH4Max ERC funding:1,992,331 €Duration:Start date: 2014-08-01, End date: 2019-07-31
European researchers have designed brain-like artificial neural networks capable of numerical and spatial cognition and written language processing without any explicit training or pre-programming. Their work, based on the machine-learning approach of generative models, significantly advances the development of self-learning artificial intelligence, while also deepening understanding of human cognition.
Research picture: © Ivilin Stoianov, Marco ZorziDetails
The research was led by Marco Zorzi at the University of Padova and funded with a starting grant from the European Research Centre (ERC). The project – GENMOD – demonstrated that it is possible to build an artificial neural network that observes the world and generates its own internal representation based on sensory data. For example, the network was able by itself to develop approximate number sense, the ability to determine basic numerical qualities, such as greater or lesser, without actually understanding the numbers themselves, just like human babies and some animals.
“We have shown that generative learning in a probabilistic framework can be a crucial step forward for developing more plausible neural network models of human cognition,” Zorzi says.
Tests on visual numerosity show the network’s capabilities, and offer insight into how the ability to judge the amount of objects in a set emerges in humans and animals without any pre-existing knowledge of numbers or arithmetic.
Much as babies develop approximate number sense without first being taught how to count, or fish can naturally tell which shoal is bigger and therefore safer to join, the GENMOD network developed the ability to discriminate between the number of objects with an accuracy matching that of skilled adults, even though it was never taught the difference between 1 and 2, programmed to count or even told what its task was.
The model was implemented in a stochastic recurrent neural network, known as a Restricted Boltzmann Machine, which simulates a basic retina-like structure that ‘observes’ the images and deeper hierarchical layers of neural nodes that sort and analyse the sensory input (what it ‘sees’).
Zorzi and his colleagues fed the self-revising network tens of thousands of images, each containing between 2and 32 randomly-arranged objects of variable sizes, and found that sensitivity to numerosity emerged in the deep neural network following unsupervised learning. In response to each image, the network strengthened or weakened connections between neurons so that its numerical acuity – or accuracy – was refined by the pattern it had just observed, independent of the total surface area of the objects, establishing that the neurons were indeed detecting numbers.
In effect, the network began to generate its own rules and learning process for estimating the number of objects in an image, following a pattern of neuronal activity that has been observed in the parietal cortex of monkeys. This is the region of the brain involved in knowledge of numbers and arithmetic, suggesting that the GENMOD model probably closely reflects how real brains work.
Learning number acuity like a child
“A six-month-old child has relatively weak approximate number sense: for example, it can tell the difference between 8 dots and 16 dots but not 8 dots and 12 dots. Discrimination ability improves throughout childhood. Our network showed similar progress in number acuity, with its ability to determine the number of objects improving over time as it observed more images,” according to Zorzi, who plans to discuss his research at the EuroScience Open Forum 2016 on 26 July in a session entitled ‘Can we simulate the human brain?’
The project’s work on numerical cognition could have important implications for neuroscience and education, such as understanding the possible causes of impaired number sense in children with dyscalculia, the effect of ageing on number skills and enhancing research into pathologies caused by brain damage.
GENMOD’s impact could be even more far-reaching in other fields, with applications in machine vision, neuroinformatics and artificial intelligence.
“Much of the previous work on modelling human cognition with artificial neural networks has been based on a supervised learning algorithm. Apart from being biologically implausible, this algorithm requires that an external teaching signal is available at each learning event and implies the dubious assumption that learning is largely discriminative,” Zorzi explains. “In contrast, generative models learn internal representations of the sensory data without any supervision or reward. That is, the sensory patterns, for example images of objects, do not need to be labelled to tell the network what has been presented as input or how it should react to it.”
A breakthrough in modelling human perception
The GENMOD team has also used deep neural networks to develop the first full-blown, realistic computational model of letter perception that learned from thousands of images of letters in a variety of fonts, styles and sizes in a completely unsupervised way. By inputting random images of natural scenes beforehand, the network learned over time to define lines, shapes and patterns. When it was subsequently given written text to observe, it applied the same processes to differentiate the letters and eventually words.
“This supports the hypothesis about how humans developed written language. There is no part of the brain evolved for reading, so therefore we use the same cognitive processes as we do for identifying objects,” Zorzi says. “The generative model approach is a major breakthrough for modelling human perception and cognition, consistent with neurobiological theories that emphasise the mixing of bottom-up and top-down interactions in the brain.”
Unsupervised learning neural networks could also be put to use for a wide variety of applications where data is uncategorised and unlabelled. For example, the network could be used to identify features of human brain activity from functional magnetic resonance imaging that would be impossible for other technology or human observers to explore. It could even be used to make smartphones truly smart, imbuing mobile devices with cognitive abilities such as intelligent observation, learning and decision-making to overcome the growing problem of network overload.
“Our findings demonstrate that generative models represent a crucial step forward. We expect our work to influence the broader cognitive modelling community and inspire other researchers to embrace the framework in future lines of research,” Zorzi says.Project details:Researcher (PI):Marco ZorziHost institution:Universita Degli Studi Di Padova, ItalyProject:Generative Models of Human Cognition, (GENMOD)ERC call:Starting Grant , ERC-2007-StG, panel SH3Max ERC funding:492,200 €Duration:Start date: 2008-06-01, End date: 2013-05-31
Marco Zorzi is a Full Professor of Cognitive Psychology and Artificial Intelligence at the University of Padua. He leads an interdisciplinary research group, the Computational Cognitive Neuroscience Lab, that explores the computational bases of cognitive functions such as numeracy, spatial recognition, visuospatial processing, reading and writing.
Imagine your favourite football team entering a stadium. An army of wireless cameras is following the players to give you the best possible view – of the whole pitch, of the chanting crowd, of each footballer, from the tip of his head to the grass blades he treads with his cleats. Thanks to Prof. Leif Oxenløwe’s research, this kind of wireless ultra-high definition television broadcasting can one day become a reality.Details
Combining cutting-edge technologies
Athletes are not the only ones hoping to set world records. With his ERC Starting Grant, Prof. Oxenløwe, from the Technical University of Denmark, has pushed the boundaries of optical communication. He is now engaged in the race to enhance wireless communications, the type that would allow high-speed video transfers and which would open a world of possibilities for extremely-high definition broadcasting.
“It is a race for higher and higher speed”, says Prof. Oxenløwe, “and we are investigating the extremes”. In 2014, his team set the world record for data transmission through an optical fibre cable with a single laser source, reaching 43 Terabit of data per second - enough to download the whole Spotify library of 30 million songs in less than a minute. Now, they are trying to leverage this record to improve wireless transmission.
“The focus of my ERC grant was to explore high-speed optical telecommunication systems with low-energy consumption. But as we talked to another team in our department, we realised the potential in combining our results with the wireless Terahertz emitter they were working on”, explains the grantee. “This is one of the positive aspects of ERC grants: the freedom to follow new, unexpected ideas.” The team patented the concept and could test the value of the prototype further thanks to an ERC Proof of Concept grant.
Their 2015 milestone, a wireless speed of 60 Gigabits per second (Gbps), or 32 times the data transmission rate needed for full HD images to reach viewers’ TV screens, was achieved by using the 400 Gigahertz (GHz) frequency range. “And now, we exploit this frequency range even further”, says Prof. Oxenløwe. At the international Opto-Electronics and Communications Conference (OECC 2016) taking place in Niigata (Japan) last week, his team will present the next step: a demonstrationof a data transmission rate of 160 Gbps in the “THz regime” (300-500 GHz) achieved in their labs. “This is the highest wirelessly transmitted data rate ever demonstrated and shows the potential of higher carrier frequencies for wireless communications.”
The future of events broadcasting
With 60 Gbps speed, a smartphone could already download a Bluray movie in less than four seconds. Unfortunately, these extremely high-speed wireless transfers are not designed for this purpose, as they require an antenna accurately set and directed to the sending source.
This technology, however, could be a breakthrough, for example, in broadcasting sport tournaments and music festivals. Nowadays, TV cameras recording live events are usually connected to a mixing centre through a cable. Prof. Oxenløwe’s prototype would allow the use of ultra-high definition cameras to send their footage to the mixing centre without the need to wire them. In the future, we could see every single centimetre of the running track, the pitch or the concert stage recorded and broadcast in real time and in unprecedented quality.
According to Prof. Oxenløwe, the technology could also become a tool for emergency teams in the event of natural disasters: “Medical and security staff could easily set up local, flexible, high-capacity mobile communication units to retrieve huge amounts of data, for example patients’ medical records, in a matter of seconds.”Project details:Researcher (PI):Leif Katsuo OxenløweHost institution:Danmarks Tekniske Universitet, DenmarkProject:Serial Optical Communications for Advanced Terabit Ethernet Systems, (SOCRATES)ERC call:Starting Grant , ERC-2009-StG, panel PE7Max ERC funding:1,518,387 €Duration:Start date: 2009-09-01, End date: 2014-08-31
Presentation reference July 2016:
“THz Photonics-Wireless Transmission of 160 Gbit/s Bitrate,” Xianbin Yu, Shi Ji, Hao Hu, Pengyu Guan, Michael Galili, Toshio Morioka, Peter U. Jepsen and Leif K. Oxenløwe. The 21st OptoElectronics and Communications Conference (OECC2016) and the conference on Photonics in Switching 2016 (PS2016), Niigata Japan, Postdeadline paper PD1-2.
Cardiovascular diseases (CVD) are a major cause of morbidity and mortality in Europe. Prevention relies on measuring traditional risk factors such as age, gender, hypertension, diabetes, hypercholesterolemia and smoking. However, many individuals, apparently at low-risk, still develop CVD. Improving predictions beyond the traditional risk factors is the challenge undertaken by Prof. Olle Melander.Details
According to recent investigations, some genetic variations are strongly linked to CVD but the mechanisms leading to the development of such conditions are largely unknown. This ERC-funded project focuses on a cohort of 60 000 unique individuals to identify these mechanisms and has already shown some promising results.
Prof. Melander and his team, drawing on previous findings, have discovered 58 gene variants linked to coronary artery disease. Thanks to this genetic information they could identify “hidden high-risk individuals” who would remain undetected considering only the traditional risk factors known today. These individuals could preventatively be treated with statins, usually used to lower cholesterol. The team has also found that a high level of neurotensin in the plasma is a strong predictor of CVD and diabetes. Neurotensin can be reduced with certain diets and drugs, decreasing the risk. In addition, they have provided evidence that higher water consumption reduces the level of vasopressin, which is associated with cardiometabolic diseases. Finally, they found that high concentrations of the amino acids tyrosine, phenylalanine and isoleucine are also a risk factor for CVD. The levels of these amino acids are high in individuals who consume large quantities of red meat and low in those with a high milk protein intake.
Prof. Melander's ultimate goal is to provide novel targets for pharmacological prevention and suggest focused lifestyle interventions. His work could have a clear impact on clinical medicine and prevention of CVD.Project details:Researcher (PI):Olle Sten MelanderHost institution:Lunds Universitet, SwedenProject:INTEGRATION OF GENOMICS AND CARDIOMETABOLIC PLASMA BIOMARKERS FOR IMPROVED PREDICTION AND PRIMARY PREVENTION OF CARDIOVASCULAR DISEASE, (CARDIOPREVENT)ERC call:Starting Grant , ERC-2011-StG, panel LS7Max ERC funding:1,500,000 €Duration:Start date: 2011-12-01, End date: 2016-11-30
Olle Melander is Professor of Internal Medicine at Lund University and a consultant at the Department of Internal Medicine of Skåne University Hospital in Malmö, Sweden. His research focuses on improvement of cardiovascular risk stratification and identification of potentially life-style and drug-modifiable mechanisms behind diabetes and cardiovascular disease.
Stronger than steel, conducting electricity better than copper and heat better than diamonds: these are some of the promises held by carbon nanomaterials. Although not as well-known as graphene, carbon nanotubes (CNTs) show these properties – offering also a great advantage: they can be produced in larger quantities. Prof. Michael De Volder now explores new ways to manufacture CNTs-based devices with optimal features, potentially opening the way to their broader commercial use.
Researcher picture © Michael De Volder
Caption: Strain engineered Carbon nanotube clover fieldDetails
While wonder material graphene was first isolated in Manchester in 2004, carbon nanotubes (CNTs) have been actively researched since the 1990s. These cylinders of one or more layers of graphene are already used in nanotechnology, electronics, optics and material science. But many engineering applications are still not possible. The difficulty comes when building larger structures such as wires, as CNTs lose their assets to a great extent.
Prof. De Volder aims to develop new technologies to assemble CNTs into organised, hierarchical superstructures that would retain their exceptional properties. His team brings together engineers, chemists, physicists and material scientists to look at the process at different scales: from material synthesis and surface chemistry at nanoscale, to the form and structure of the material at microscale, up to the larger scale: how to integrate CNTs ‘building blocks’ into 3D all-carbon devices.
The team has already managed to produce microstructured surfaces with tuneable characteristics such as stiffness or strength. To do so, they provoke the CNTs bending as they grow, managing to form controllable complex shapes in 3D. These compound surfaces could also replicate the water-repellent or adhesive features found in the skins of certain plants or animals for example.
Developed in the HIENA project, the technique is already used to make biomimetic smart surfaces but also chemical microsensors and batteries. Prof. De Volder’s team also works on CNTs application for extremely accurate water purifiers and more efficient energy storage systems for electric cars.Project details:Researcher (PI):Michael Franciscus Lucas De VolderHost institution:The Chancellor, Masters And Scholars Of The University Of Cambridge, United KingdomProject:Hierarchical Carbon Nanomaterials, (HIENA)ERC call:Starting Grant , ERC-2013-StG, panel PE8Max ERC funding:1,496,379 €Duration:Start date: 2014-01-01, End date: 2018-12-31
Dr Michael De Volder is the principal investigator of NanoManufacturing Group at the Department of Engineering of the University of Cambridge. He carried out his PhD research in Belgium and Japan. He then joined the Massachusetts Institute of Technology, the University of Michigan, and Harvard University as a postdoc researcher. He is a Laureate of the Belgian Royal Academy, and vice president of a nanotech start-up company.
While computers can calculate or recognise faces, they are not aware of themselves (yet?). Consciousness is in the essence of human beings; its nature, however, appears to lack a reliable explanation. Prof. Axel Cleeremans is developing a new theory, the Radical Plasticity Thesis, maintaining that consciousness is a long-lasting property of our brain rather than just a static feature. In order to test it, he is taking a multidisciplinary approach including psychological studies and advanced brain imaging.
Image: ULB — CRCN Portrait: © F.R.S.-FNRS — Jean-Michel BYLDetails
Our consciousness, or the movie of our own lives, is probably the main feature that makes us different from machines. The Radical Plasticity Thesis, a new theory for understanding our mind, states that consciousness is something that the brain learns to do rather than being a static property of our brain.
We are continuously trying to predict the consequences of our actions in order to minimize surprise. For this purpose, our brain is always reinventing its theory about itself, so developing internal models shaped by its experience interacting with the world, with other agents, and, crucially, with itself, Prof. Cleeremans affirms. According to him, we are constantly learning, consciously and unconsciously, to improve that story. If the Radical Plasticity Thesis is demonstrated, our own ability, as well as our brain's ability to learn to better re-define our knowledge would be what make us conscious.
For Prof. Cleeremans, consciousness depends first on the quality of the representations we continuously create as we interact with the world and with other people: through learning and attention, strong, stable and distinctive representations emerge and can then become redescribed in such a way as to become conscious representations. Being conscious, in this sense, always involves knowing that we know: the brain is looking at itself, in the service of better control of action.
To prove his theory, Prof. Cleeremans’ team is conducting a multidisciplinary study which mainly consists of behavioural experimentation, examining how people react to different types of events and experiences. These experiments will be complemented by computer simulations and brain imaging, using innovative techniques such as virtual reality to map human consciousness.Project details:Researcher (PI):Axel Noël F. CleeremansHost institution:Universite Libre De Bruxelles, BelgiumProject:The Radical Plasticity Thesis: How we learn to be conscious, (RADICAL)ERC call:Advanced Grant , ERC-2013-ADG, panel SH4Max ERC funding:2,286,316 €Duration:Start date: 2014-05-01, End date: 2019-04-30
Axel Cleeremans is a professor of cognitive psychology at the Université Libre de Bruxelles, where he directs the Centre for Research in Cognition and Neurosciences. He is also a research director at the Fonds de la Recherche Scientifique. His research centres on the differences between information processing with and without consciousness, particularly in the domain of learning and memory. In 2015, he received Ernest-John Solvay Prize for Human Sciences.