Say Hello to Tomorrow's Tech
Bonjour! This week we’ve been to Paris where we’ve been attending “Hello Tomorrow”, the summit that showcases world-changing emerging technologies that are about to make it big. This week: A tiny microphone that lets you zoom in on individual voices in a conversation, the perks of dissolving your mobile phone, and from kites to kilowatts: a new form of wind power...
In this episode
01:16 - Finding one voice in a crowd
Finding one voice in a crowd
with Sahil Gupta, Soundskrit
The market for voice activated products is exploding. Millions of people are embracing this technology. Many of you probably use it yourself: Amazon have their Echo, and rivals Google have their home assistant, and there are numerous other new recent entries to the market. But all of these voice-activated gadgets face a similar challenge: they struggle to pick out the person talking to them in the cacophony of a noisy environment. At the moment, they solve the problem by including a bulky array of microphones that resolve the direction of a sound and allow the device to decipher what’s being said and by whom. But this takes up a lot of space and ultimately constrains the design of the product. Now a company called Soundskrit reckon they have the answer. Taking their inspiration from chirping insects, they’ve been able to shrink a microphone to something smaller than the nail on your little finger. Adam Murphy heard how, from Soundskrit’s co-founder Sahil Gupta...
Sahil - We're working on new type of directional microphone to improve audio capture and speech recognition in noisy environments. You know, we've seen a lot of different voice-based applications emerge that require the ability to sense sound from a distance. And the challenge is, you have a lot of noise in the world so when you're trying to pick up the voice of somebody that standing a few feet in front of you, you also pickup all the background noise. And to combat that people usually have to use very big microphones or microphone arrays and that concept can't really be shrunken down well, so we have a new directional microphone concept where we can kind of create a single chip solution smaller than my fingernail and give you very good directional performance and noise rejection.
Adam - So how actually does it work because it would with very useful here in a very nosy conference centre?
Sahil - Yeah, yeah absolutely! So it's interesting right because the concept was originally inspired by insects actually. We’re working with professor who has been studying insects for the last 30 years, because they have a really small auditory system, and they use the hairs on their body to send sound. Imagine I had a string that I held between my two fingers, if I blow across that string it’s going to move, but if I blow parallel to it, it won't move right. And so it only senses sound, air, coming from a specific direction, and I can take that same string and layer multiple strings in different directions and now I can kind of mechanically filter out the sounds coming from each different direction, and then in software I can combine anything and mix it together or where I can only listen to a specific one depending on the application.
Adam - And how good have you found it to be so far?
Sahil - It works quite well. I mean in this sort of environment you know we can have a couple of people talking into the microphone at the same time and we can transcribe what both of the people are saying. We can localise where you're standing from, where you're moving, where you walking, even with all the noise. It’s still very early proof of concept and we definitely have a lot of R&D before it’s ready for the market. But yeah, right now it's looking good.
Adam - What do you hope the end product looks like?
Sahil - it would be just a single sort of, chip that you could fit into something like your phone or your Apple Airpods, or your smart speakers and smart home devices. That's on the hardware side and there we’re also developing a sort of software layer. So we give you this directional microphone, we also want to provide a sort of toolset where you can have new features you can mess around with to get different types of effects and whatnot.
Adam - What kind of things are you envisioning people do with that?
Sahil - You know there's obviously speech recognition right. So think about the Amazon Echo, it's this big, large clunky speaker, we could give you that type of capability on just a single chip that you could drop into any of your devices. And outside of speech recognition right now with my phone if I pinch my two fingers I can zoom in and out of my video, what if I could do that with my audio as well. Now I can zoom in and out of my audio and even filter out more of the background or take more of it in, or I could tap on my iPhone and wherever I touch on the screen it will kind of selectively only listen to the sound in that portion of your picture and image. Also spatial audio, right now there's been so many people looking at binaural playback, stereo playback from headphones. Where you can get the content from, right. Now from your smartphone you can actually capture that spatial content and then when you play it back through a headset or speaker you have the content to actually play that back.
Adam - So it's loads of little microphones working together, is that what the strings are basically?
Sahil - Yeah, yeah pretty much. And the nice thing is we use these nanostructures that we can package on a single chip so to the end customer it looks just like a normal single microphone. And then you know if you want to do some really crazy stuff we could use multiple of our microphones and get even better nose rejection, even farther listening ranges, things like that.
05:47 - Noise-cancelling windows
with Aman Jinal, DeNoize
Noise is frequently cited as one of the leading causes of stress. People living near airports and busy roads complain that they can never relax because of intrusive sounds. Their experiences are reflected in higher levels of mental ill-health, and physical problems like high blood pressure, strokes and heart attacks. Chris Smith spoke with Aman Jindal from the company De-Noize, who might have a better solution...
Aman - In any building, 70 to 80% of your noise enters through your glass because they're the weakest link. And as you mentioned about double glazing, yes it's better than a single glazed window for sure, but they only work in very high frequency regions. When you have this buzzing sound of a tyre on a highway or of planes, the rumbling sound of the engine, those are the low-frequency sounds and they are not stopped by any glass unless the glass is a metre thick and that's just not feasible on windows.
Chris - So when sound hits the house it's basically a pressure wave, isn't it? It's hitting the glass surface and it's pushing the glass in and out and that's making the inside surface of the glass move in and out, so it's basically transmitting that sound into your room, which is why we're disturbed by jets overhead and cars on the road?
Aman - Exactly correct. And if we can cancel these vibrations already on the glass, then you do not hear anything.
Chris - You're trying to make noise cancelling headphones for windows?
Aman - In a sense, yes. Concept wise we do the same thing. We vibrate the window in an opposite phase when we know how it's going to vibrate, and that cancels everything. In essence, it's a very simple thing, whenever we know the centre of the window should be moving towards on the inside of the building, we move it towards the outside of the building and then it cancels each other.
Chris - So you must have some kind of microphone array or some kind of system that is looking at how the window is trying to move when the sound waves hit it, and you're doing some clever processing to work out what force you'd have to put into the glass in the opposite direction so that when the glass tries to move from the sound it doesn't, and then we end up with basically no movement inside so we get basically a quiet room?
Aman - Exactly! The sensors or the actuators that we put are put on the side of the window, so whenever the window vibrates we sense it on its edge and then we input the forces from its edge to vibrate in an opposite way to achieve the cancellation. Those actuators and those sensors, they are combined so they can switch their role so they can at points act like a sensor, and at point act like an actuator.
Chris - How good is it?
Aman - It can achieve up to 90 to 95% additional reduction. On a scale of how the noise is measured, it's measured in a unit called decibel which is a logarithmic scale, we can achieve up to 30 decibel reduction. So 30 decibel is actually a really big reduction, a game changer in this aspect.
Chris - And will it work across-the-board? So does it matter whether it is a chainsaw in the garden, a car going down the road, or a jumbo jet coming into land at Heathrow? It doesn't matter what the sound source is it will still work or are they some things you just can't get rid of?
Aman - A lot of noises are actually stopped by the glass itself, because of a physical barrier. The only thing that's not stopped is these low-frequency noises coming from the jet engine or something like that. And that's the one you hear because that's the weakest thing and we focus on cancelling those, so once they are cancelled you wouldn't hear anything.
Chris - Do you need special glass for this or could you retrofit your system to existing glass? Because glass is expensive.
Aman - It doesn't matter which glass you use, but at the moment how far the technology is it cannot be retrofitted because it's a very complicated process of putting all the sensors and actuators in a specific way.
How we see this on existing windows, people can upgrade it by calling a professional who knows how to put these sensors and actuators in there and then it would work. But we do see our launching market as a new window, a new real estate market where the windows or glass structures would come already fitted with our technology.
Chris - Is it expensive to run? How much energy is going into cancelling out the sound because you've got to put in as much energy that is hitting the building to cancel the sound, haven't you, so it must be consuming quite a bit of energy?
Aman - Our actuators actually run on a very small energy. To give you an example, for a metre square of window you might be spending 10 to 12 watts, which is not a lot of energy.
Chris - And what about price?
Aman - I cannot give you an exact number at the moment, but I can give you a ballpark. We estimate that for a metre square window, which on an average in the UK costs around £500 at the moment a double glazed window, to the end user it might be £650 or £700, but not more than that.
Chris - And how long before this is going to hit the market?
Aman - 14 to 15 months is what where estimating at the moment. We're still developing the tech a bit more and we see that in the next year.
11:59 - A simple plastic scanner
A simple plastic scanner
with Martin Holicky, Matoha Ultrascience
Recycling is becoming more important, both to ensure that we don’t waste materials but also to ensure that we don’t end up with waste where it shouldn’t be, like plastics getting into the ocean. But one issue recyclers face is dealing with all the different kinds of plastics we have. A bin bag is made from a quite different material from a fizzy drink bottle, so they have to be recycled separately. But if you don’t know what things are made of, how do you begin? In the developed world, we have big plants that can sort through all our rubbish, but in the developing world, they need a quick, straightforward way to tell one plastic from another. And tech startup Matoha Ultrascience has built a portable machine that makes this sorting process quick and simple. Adam Murphy heard how it works from co-inventor Martin Holicky...
Martin - Our machine has infrared lamps and infrared optics in there and every material has a unique infrared signature. So we just put the item of plastic or textiles in it and within two seconds it tells you immediately what material it is, so far example if it's cotton, polyester or similar materials.
Adam - How does it get that fingerprint? What happens to the infrared light to generate this signal?
Martin - Infrared light interacts with the material so it gets reflected off the material and absorbed by the material. And then our optics look at which parts of the infrared light have been absorbed and from that we can deduce what the material is.
Adam - And where do you see this working? What do you think the applications are?
Martin - It's really important for any kind of recycling of both fabrics and plastics to know what's the material composition. So you need before any kind of recycling you need to sort the waste coming in. So our machine could clearly be useful for sorting of plastics, all kinds of waste so that you can then recycle them.
Adam - How specific is it? How many different kinds of plastics can you identify?
Martin - We can identify all the major types; polypropylene, PET, polystyrene, all the major types and the accuracy is quite good actually. You just put it there and it can immediately tell you what it is.
Adam - Are you hoping to go on someone’s countertop or is it for other locations?
Martin - Currently in the developed world they have this super expensive big machines which can sort the waste automatically, but because it's so expensive in the developing countries they can't afford them. So what we envisage is that we give the workers in the developing countries almost an additional pair of eyes. It can tell you what the material is and then they can sort it properly.
Adam - Now some plastics are see-through and others are not, like polystyrene is solid white, how do you deal with the two different kinds?
Martin - So we have actually two lamps in there – one shines from the bottom, one shines through the sample and depending on if the material is transparent or not the lamp of the particular kind makes the infrared light go through or reflects off the sample.
Adam - How easy is it to use?
Martin - Well it's super easy. You just put it there - you don't need a degree in chemistry or physics or anything like that. Just put it there and it tells you immediately what it is. It's like a supermarket checkout - beep, beep, beep.
Adam - And is this ready to roll out, or are you at the prototype stage or how far down the line are you?
Martin - We have been working on this for the past two years and currently we are doing the first field trials in Europe. We are nearly there and hopefully later this year we'll be able to start selling the machine.
15:25 - Wound-healing bacteria
with Evelina Vågesjö, Ilya Pharma
It sounds counterintuitive to rub bacteria into a wound to make it heal faster, but a Swedish startup, called Ilya Pharma are doing just that, with astonishing results. A few years ago, during her PhD, Evelina Vågesjö discovered a family of immune signals made by white blood cells that lure other immune cells into wound sites where they can promote repair. But these signals are tricky to produce, and they act for only a very short time. Her solution was to add the gene used in the body to make the immune signal to bacteria like those you find in yoghurt and apply the modified bacteria into the injury…
Evelina - During my PhD I investigated different molecules that immune cells use to talk to each other so that they know how to move around in the body, and especially if we get an injury I specialised in a molecule that is increased in the concentration at the injured site. So what we could actually do is to boost the process at an earlier stage and then the whole healing is faster and this we have described in animals - mice and mini pigs but we weren't quite happy with that and we want to find out if this also works in humans, so that's the next step. So we're going to do a trial in healthy volunteers is the next step.
Chris - What is the class of molecules that you've discovered that do this?
Evelina - They are called chemokines. Essentially what they do is tell immune cells how to move in the body. So they build up gradients from a high concentration to a low concentration and then the immune cells move along this gradient.
Chris - So it's a bit like the immune cells are smelling this trail, like a breadcrumb I suppose - Hansel and Gretel breadcrumb trail of where the wound is and they sniff out the wound and flock in, and you're saying if I add extra of this molecule really early into the wound I can get many more immune cells there much more promptly so I can kickstart the wound healing?
Evelina - Yeah, that's exactly how I would describe it. By forcing the wound to secrete more of this we can get much more efficient immune cells to come there earlier and accelerate the healing process.
Chris - But these chemicals, these chemokines, they're complicated, they're proteins they're not just something that you can knockout easily so how are you making the wound make that?
Evelina - These chemokines, they have a very short half life so if you are just taking the chemokine and add it to the wound it will be degraded very fast. So we have overcome this by asking a type of lactic acid bacteria to produce this for us in the wound. We took normal probiotic bacteria found in yoghurt and then we inserted a gene into them that express this human chemokine, and then when they are in the wound they produce this chemokine for about one hour to the wound's surface and that is enough to activate the immune cells.
Chris - So you're saying you put extra bacteria that you've genetically modified into the patient's wound?
Evelina - Yes we do. And it works fantastic.
Chris - So the bacteria don't make the wound worse, they don't cause more inflammation? They're churning out this immune signal that they don't worsen the wound by being there?
Evelina - No. So what was seen in our experiments is that it is quite tough for them to produces this chemokine. It takes a lot of their energy so therefore they will not live for very long. And also we use a type of bacteria, their normal environment is not the wound, so they would also be killed by the wound environment. So we find them in the wounds for about one hour.
Chris - So they don't hang around for very long because that's sort of reassuring then. Although your putting bacteria in they do the right stuff and then they're gone?
Evelina - Yeah. And during this hour they produce enough of this chemokine to affect the immune cells.
Chris - When you then look at wounds that are treated this way, what is the difference in outcome between wounds that you do colonise, albeit temporarily with these microbes, and wounds that are just left alone?
Evelina - We see that we get a lot of granulation tissue, which is the red healing tissue in the wounds at much earlier stages. We also see that we get the top part of the skin that cover the wounds fully at earlier stages. And then we also see reduce scarring and we have validated this in mice, mini pigs but also it also using skin biopsies.
Chris - One thing that's worrying me slightly is that you are putting these bacteria, which are genetically modified into a wound and they then could escape out into the environment, so is there not a risk? And I presume you've had to reassure various regulators that there is no danger or threat to the environment through the escape of these genetically modified bacteria, not just genetically modified, they are making human immune signals?
Evelina - This is something that we are following. We don't see that they have spread to the environment but this is something that we will continue to investigate and this is, as you say, we are obliged to do this by regulatory bodies. We have though done tons of risk assessments and we have also put them out in the parking lot and tried to find them and so on. But we also tested in - we took away all my water pipes in my bathroom because you can't just test on just new pipes - I mean that's not the real environment. Basically the worst-case scenario with the bacteria that we are using it was isolated actually from the rat intestine. We have other pilot data in the model of inflammatory bowel disease. So if they would spread to the environment they cannot really survive anywhere else than in rodent intestines, so if rodents would have problems with colitis, which is inflamed intestine, they would have a little bit less problem with that.
Chris - And how is your plumbing now?
Evelina - Well I got new pipes, but I was also very disturbed by how dirty my own pipes were!
21:53 - Turning carbon dioxide into fish food
Turning carbon dioxide into fish food
with Peter Rowe, Deep Branch Biotechnology
With 7.2 billion people on Earth, and rising, the carbon footprint of the human population runs to tens of billions of tonnes of CO2 every year. Now a company called Deep Branch, spawned by entrepreneurs who studied at the University of Nottingham, think they’ve got a way to harness the CO2 chucked out of the chimneys of cement factories. By feeding it to a special population of bacteria, they can turn it into fish food! Adam Murphy heard how from one of the brains behind the operation, Pete Rowe...
Pete - So it's a microbial process called gas fermentation. We can take CO2 directly from industrial sources, let's say a cement plant or a power plant that has a large CO2 output, we combine that with hydrogen, we bubble it through a liquid medium - just kind of like brewing beer but rather than producing ethanol we produce protein - and that protein is used as an animal feed.
Adam - And where do you get the raw materials to do that with?
Pete - The CO2 we get directly from an industrial partner. So at the moment we're at kind of like a pilot stage so we've built a shipping container where we've installed a scaled-down version of our equipment, put that on-site at a cement plant and installed the CO2 pipe directly from their chimney. We provide some hydrogen from a cylinder in this instance and then all we need is a bit of water, some salts, some nitrogen and that's it.
Adam - And what are the bacteria doing - the microbes sorry, what are the microbes doing?
Pete - So the microbes they are bacteria. They're thought to be the oldest evolved thing known to man. So they're really the start of life, and what they do is they use the CO2 as a carbon source and they use the hydrogen as an energy source. So if you think about a conventional fermentation process that might use sugar, let's say you were making wine, you'd be using grapes that have got a lot of sugar in them, and the yeast in that process uses the sugar as energy to grow. And it has carbon in there and they need a carbon source to build the molecules of life.
Now we have a more stripped down version with our bacteria from ancient times when sugars weren't really available so instead they used the CO2 as a carbon source, as this molecular building block for life, and they use the hydrogen which is very energy rich as an energy source.
Adam - What is the protein, is it the ground up bacteria or what is the protein which you make?
Pete - The product is a biomass. So the same way in which you think about soya bean being a good protein source, it's about 30% protein in a soya bean, but what you're feeding to people or to animals in a soya meal is just a ground up version of the soya bean. So with us it's a biomass as well, it's the bacterial cells that form a pulp and within that we have 60% or 70% protein, so the protein yield is really good. And, of course, if you think about how much water, how much fertiliser it takes to make soya beans, or in the instance of fishmeal how much fish you have to catch to feed animals, with us it's a lot less resource intensive and, of course, if we’re capturing the carbon it's got a sustainability angle on there as well.
Adam - And the protein, are you feeding it to people or what's your approach to use it?
Pete - I mention fishmeal, this is a real problem in aquaculture so fish farming. Soy meal isn't really appropriate there because the fish can't digest soya very well so instead fishmeal is used and this is fish that's caught from the sea. Of course, there's a finite amount of fish there. People want to eat the fish, fish farms want to use it as a feedstock and fish stocks are dwindling, prices are going up so the aquaculture sector is actively looking for other protein sources. So they're looking at insects for instance, but they're quite hard to scale; whereas our technology is only limited by the amount of CO2, and if you think about a cement plant that's not far off a million tons of CO2 per year, so there's plenty of that going around.
Adam - It sounds like a "two birds with one stone" kind of job, taking concrete and turning it into fish food?
Pete - Precisely, yeah. So sustainability on two sides, yeah.
Adam - And what stage you are at now and what you looking to do?
Pete - Yeah. So now we have full lab-scale validation and I spoke a bit earlier about this mobile production unit that were going to deploy with our partner who is a cement manufacturer. Once we get full validation that it works with industrial gases rather than just CO2 in a lab, we are then looking to scale up to a pilot plant that will produce tonnes and at that stage we can start getting full validation that it works as a fish feed in this instance. And then within three or four years we hope to reach full commercial scale, whereby you'll be able to buy aquaculture products, so fish on the supermarket shelf, that are fed with our protein.
Adam - Have you fed any fish with it yet?
Pete - We've done a few tests in the fish tank, but not any commercially relevant species because our production volumes aren't enough to get good data back on that yet.
Adam - Well do your fish tank fish seem to like it?
Pete - Yeah, they love it!
Behind Hello Tomorrow
with Sarah Pedroza, Arnaud de la Tour, Hello Tomorrow
“Hello Tomorrow” as a concept actually began in 2011. It was born out of the frustration of a handful of early career scientists who knew they had world-changing technologies on their hands but they had no easy way to commercialise them. Eight years later, thousands of start-ups now bid for a chance to be among the golden few chosen to present and pitch to mainstream investors at the conference. Chris Smith went behind the scenes, to talk to Sarah Pedroza and Arnaud de la Tour, two of the key people who make it happen...
Arnaud - I'm Arnaud de la Tour, CEO of Hello Tomorrow…
Sarah - ...And I'm Sarah Pedroza and I'm the co-Managing Director of Hello Tomorrow.
Arnaud - We were doing our PhDs and we were a bit frustrated because all the money, all the hype, was going to digital platforms; what is the next Uber of anything? And we were seeing so many great inventions and technologies in the lab but we needed the mindsets, the business, to pay attention to this. So that's why we created Hello Tomorrow and the conference and the start-up competition.
Chris - How long's it been going?
Sarah - Arnaud funded Hello Tomorrow in 2011 and I remember it quite well. And Demis Hassabis was on stage, it was actually the actual CEO of Deep Mind and nobody knew him.
Chris - I know Demis - we've had him on our programme - because he was at Queens' College, in Cambridge, where I am. He came on the programme because he went on to do a PhD in neuroscience after his degree and used his computer knowhow to solve a lot of neuroscience problems on the brain and then went back and founded Deep Mind. So you got him!?
Sarah - Yes we got him! He wasn't that well-known at this time and I was absolutely amazed to be honest. I was one of the volunteers at this time and I thought, okay so if these bunch of students and volunteers comprising of Hello Tomorrow is able to bring these brilliant minds it has a very strong potential.
Chris - I've been watching this morning a succession of extremely good pitches because you're bringing in companies which are nascent and they're really early technology and you've brought an enormous number of them to Paris. They're presenting what they have in mind, the problem of trying to solve, how they're going to solve it and why someone should invest in it. I've never seen anything like this in the sort of volume that you've got going through here. Where do you find all these companies and how do you get them here?
Arnaud - We have so many startups coming because we establish lots of relations with universities, incubators, accelerators all over the world; America and North America and Europe and Asia. Everywhere! At the beginning, the reason why they applied was because we were giving them a prize money of 100 000. But today the reason why they come is because they get some visibility. We got 4,500 applications and the 500 best, they are called the top 500, and people contact them immediately when they get the award. It's life changing for the company. It's more visibility and definitely also the connections. Yesterday, we organised what we call the “Investor Day”. We had 150 investors and almost 300 startups and they had more than 1000 meetings in the whole days. And for them meeting this crowd of people who understand their projects the technology behind it. It's quite unique. That's why they come from all over the world for this event.
Chris - Having done this for a little while now, have you got evidence that this is leading to investment?
Sarah - Yes. Actually the first winner we had, I remember Grégoire Courtine pitching on stage as a spinoff from EPFL. He found his CEO at the Hello Tomorrow global summit which means that he found the person who was able to raise funding and to help him out and they - if I'm not wrong - raised more than 35 million after the summit. We also have Lilium Aviation who was the winner two years ago and basically they raised more than 100 million after. So I think we have a good track record. But let's be clear, we help, we are the right platform but they are the amazing projects. So let's let's not take too much credit.
Chris - Indeed but are you unique? Is anyone else doing this and the way you're doing it?
Sarah - I don't think so. I think a lot of organisations are positioning themselves to do so. The thing is that Hello Tomorrow has been funded by science entrepreneurs. We understand them and they feel that they are into the right community. They find their own tribe and this is something that’s quite unique, I think, about what we do.
A helping hand with walking
with Alex Sancho, MyoSwiss
The ability to move about is something most of us take for granted. But, for many people, mobility is not a simple issue, and they could really use assistance walking. Myoswiss have designed the “myosuit”, which works a little like an electric bicycle for your legs. A lightweight battery backpack is connected to a frame that wraps around each leg. When you take a step, Myosuit uses a system of metal strings at the knee which stretch and pull your leg, making the same natural movements that you leg muscles do. The idea is to give a helping hand (or leg) to anyone going through rehabilitation, or even just struggling to climb stairs. Alex Sancho showed Adam Murphy how it works...
Alex - The Myosuit is basically an exo-muscle and what an exo-muscle does is, we provide an extra layer of assistance for patients that have some difficulty while they're walking. So be it elderly patients or patients that have a muscle dystrophy or incomplete spinal cord injury. And what we do is support them during their daily lives. So we can give support while they stand up, while they walk, while they go upstairs.
Adam - So you're here wearing what looks like a fancy robot suit. When you try to move your legs, what actually happens what goes on mechanically?
Alex - One of the beauties of our system is that compared to other similar systems which can be an exoskeleton for example is that the system is really really simple. So we have a backpack that's four and a half kilos. The backpack has two motors with two tendons attached them. The tendons are very smartly routed in such a way that when you pull on them, both your knee and your hip extend. That's a movement that you do when you stand up from a chair. On top of that we have a sensor layer that has two IMU sensors, so a gyroscope and an accelerometer per leg. And also we have one on the trunk. So what that does is we are real time creating a picture of the person wearing the suit and based on that how we can parametrise the gait cycle of the patient or the activity that you're doing and apply the force that supports them at any given point. So we are lighter and we can assist through many more activities, very dynamic because since we only have one motor that is pulling on a cable we can have very dynamic responses to the movements of the patients.
Adam - So instead of forcing the joints itself to move it pulls on tendons like exist in your body now.
Alex - Exactly, so we are always trying to mimic our own bodies. Also we have a passive layer so the tendons when pulled by the motors they extend the muscles but also we need a flexor which would assist when you are swinging your leg forward. So what we do is we assist with the motors when you put your feet on the floor. Basically give support during that time and then you move your leg backward and at some point the passive elements which are springs sort of, they get tension and the motor releases and your leg gets propulsed forward by the springs. So the beauty is that with a very simple system that actually mimics the body and the muscle structure. We are actually supporting throughout the whole gait cycle.
Adam - How does it know when to stop when you put your foot in the ground and it doesn't keep extending, how does it know you've taken a step?
Alex - So our algorithms, what they do is when you put the feet on the floor. If you look at the signals from the IMUs you get very similar signals regardless of the patient or even if it's a healthy subject. So we detect that exact moment and basically the actuation is very simple, and for the release of that what we do is ask the patients to calibrate so they walk for six steps and based on that we get more or less the way they walk. And based on that we can estimate when the suit should release the force.
Adam - Are people using this now to help them?
Alex - Yes. So what we are doing right now is a feasibility test, to test on patients and to define the usability. At the same time the patients are very happy to test it and experience the technology. So up to now we have tested on around 17 patients so incomplete spinal cord injuries, myopathies, a couple of them stroke patients. One of the virtues that the suit has is that we can assist only one leg instead of both legs. So with stroke patients that have severe damage on one side we can assist on that side and leave the other side transparent, we call it, which is when we don't assist the patient.
Adam - And how easy is it to wear and put on because I've seen you've been wearing this most of the day.
Alex - I have to say that it's comfortable like it's wearing around a 4 kilogram backpack and we are targeting lower weights for the commercial version that will show up at the end of the year. Also the knee support; we are aiming at also reducing the size. So yeah in general we want to reduce the weight significantly and also the donning time of the suit to like two minutes which would be way faster than anything on the market.
Dissolving your phone
with Euan Doidge, Imperial College London
They say that silence is golden, but so, it turns out, are our mobile phones. In fact, they’re stuffed full of rare metals. And since we’re beginning to run low on our supplies of these materials, and demand for technology is rising, how can we get the precious metals back out once we don’t want the phone any more? One way might be to chuck it in a vat of acid! Eva Higginbotham spoke with Euan Doidge, from Imperial College London, who’s looking at how to do exactly that...
Euan - All the metals in the world mainly come from rocks. The problem is there's not a lot of metals in those rocks. It takes a lot of energy and effort and resources to get not a lot of metal out, for a lot of input in the first place. Our solution is to try and recycle as much metal as possible from secondary sources, such as mobile phones. So whereas you can only get up to a gram of gold out from a rock, you can get 300 grams out from a mobile phone.
Eva - I never realised that my phone contained so much gold. What is the gold doing in there in the first place?
Euan - The gold is present in your printed circuit boards as the contacts between components. So it's not exclusive to mobile phones, it's anything with a circuit board will have gold present. It's not there really doing anything but being an electrical contact. I should say there's not a lot of gold in any one person's phone - it's only 30 mg worth, 85p worth. It's when you start getting lots of phones together that's when it becomes much more viable and important.
Eva - So how are you trying to extract the gold from the phone?
Euan - The classic way of doing this, the kind of low-tech way is pyrometallurgy, essentially melting metals out of the source. That's not green, it's not sustainable and it's not really great just to have more emissions going into the atmosphere. Our solution is not green, it's greener, it's more sustainable. It's a technique called solvent extraction. It was first pioneered during the Manhattan project but we our using it now to recover metals in solution. So rather than using high energies we using just liquids. Essentially we're dissolving all your metals into one aqueous solution, designing something that's selected for one metal and extracting it into an oil. Once you do that you can separate off the oil, recover the metal that way and you've got selective and efficient recovery without using high temperatures.
Eva - Cool. And so how long does that process take?
Euan - It depends. Everything's got a rate determining step somewhere. The first step might be how long does it take to dissolve your phone in the acid, then the separation stage needs to be quick. We aim to do that in two minutes or less but this happens in a continuous flow. So that's from the very start to the very end that can take ten minutes but any one section should take no more than two. The reduction step again, dependent on time, how quickly do you want to do it? The quicker you do it the lower purity you get, so it's how long do you want it to be essentially for the quality of products you want.
Eva - So it goes from the phone, I guess you might strip away some of the screen and stuff, those sorts of parts first and then you just put the circuit board into the solution?
Euan - So this is the big question: what's the most efficient way to dismantle the phone? So our solution is to dissolve the whole circuit board in an acid solution, that gives you one big mixture of metals, then you selectively take out each metal. Now the choice is, do you pre-dismantle the phone. Do you separate as many components as you can to make it a simpler mixture or do you do selective leaching. So before you dissolve it do you say well I'm only going to try dissolving some of the metals and then separate those. My preferred option is to dissolve the whole phone.You've gone to the effort of dissolving it; then get the metals out; that's one step rather than multiple.
Eva - And about what sort of percentage of recovery are you getting?
Euan - That's an interesting point. So any one cycle will load a hundred percent. Now there's always going to be loss and you have to do things in sequence. So as something goes forward and then something goes back the other way, but together you can recover all of the gold. It's a remarkably efficient process.
Eva - You work on this in an academic context as well, is there a company that is trying to do this as well?
Euan - Lots of companies are now thinking about this. They've realised there's only thirty years left of copper, thirteen years left of indium and potentially we've already run out of lanthanides. We need to start recycling metals. So all the big mobile phone companies, tech companies are now thinking about how do we recover these metals.
Eva - And once the gold or whatever other metal you're taking out of the phone is extracted, can it just immediately be repurposed for something else?
Euan - So this is the big question about how you want to recover the gold? Everyone's going to want it in slightly different form, so our proposal is we give you a pure solution of gold and then you can choose to reduce it, to replace it or you can just precipitate it and get lumps of gold that you can melt down to make rings.
Turning kites into kilowatts
with Edgar van Nunen, Skypull
The UK gets almost 30% of its energy from renewable sources. But that means that over two thirds still come from non-renewables, like coal, oil and gas. And the signs are that we need to improve our act quickly, because time’s running out to cut carbon emissions. So a new way to harvest energy from the wind sounds very attractive. What Skypull have invented is a system that uses a kite - equipped with drone motors to enable it to be controlled - which flies much higher than the average wind turbine; and by pulling on its rope, it generates large amounts of electricity on the ground. Chris Smith spoke to Edgar van Nunen, part of the Skypull team, to learn more…
Edgar - We want to make a difference to energy transition to renewable energy to replace fossil fuel energy use, and that is a big problem. Fossil fuel use is growing twice faster than renewable energy production.
Chris - So what's your solution?
Edgar - We are developing a system producing more energy but also doing that in more locations. Our system works with a drone that operates fully autonomously at altitudes up to five times higher than wind turbines reach.
Chris - So we must be talking one hundreds of metres half a kilometre to a kilometre up then
Edgar - About until 600 metres, is that what we're planning. With that we're staying under the altitude of commercial air traffic but at the higher altitude there is far more wind available across all of Europe.
Chris - So explain to me how this works then you're saying it's a drone aircraft that's going to generate electricity from wind. How?
Edgar - Correct. So it's a drone it takes off from its platform on the ground station which has the generator so we have two elements. One is the drone and the other a ground station with generator. They're connected by a tether or a rope if you like. And it functions as a kite. So it goes up with motors but done with minimum wind speeds to sustain the drone in the air. Maybe you know when you were little; the kite stays in the air you can steer it and actually when you take a little bit bigger kite like some people do like in sport kiting it becomes quite a bit of force and you feel it in your arms. Now if you make that a little bit bigger you can actually say hey that force that I feel I can do something with that energy and that's what we do. We harvest the kinetic energy of that wind and it is exercised on the tether which is connected to a winch on the ground which in itself is connected to a generator. So by going up we are generating electricity from lift force of our drone which is operating without motors of course which are pulling the tether and driving the generator on the ground.
Chris - This is ingenious so basically the thing ascends. It's being pulled up by the wind and you let the cable out which is generating electricity on the way up. Then what you put it into a dive so it loses altitude again and you reel it in because it's obviously not going to take an energy to reel in the slack tether as it comes down and then you just keep repeating this in cycles up and down and every time it's on an up cycle it's pulling hard on the string and that's generating electricity.
Edgar - Exactly that's how it works. We call it the yo yo cycle so like a yo yo it goes up and down. We can also like you do in sailing you move it partly out of the wind then you're only let's say taking part of that power from the wind, so you get into a range where from a low wind speed, enough to sustain the drone in the air, to a very high wind speed can be handled by the system.
Chris - How big is this thing?
Edgar - So our current prototypes are very small still that is proving the concept. So we're talking about one metre wingspan and kilowatt scale production.
Chris - So despite only being a one meter across kite effectively that's generating electricity at the rate of a kilowatt?
Edgar - Potentially up to four kilowatts.
Chris - That's a lot.
Edgar - It is. And the next level that we want to develop to really demonstrate in the market that this works, that it is efficient, and that it is safe, which is a drone plan at a six meter wingspan. Now a 6 meter wingspan, we talk about a 60 kilowatt system. The ultimate plan is to go to a megawatt scale.
Chris - Will this be deployed as, in the same way as we have a wind farm, or we have a solar farm with a big array of generating entities. Would you have an area of the country mapped out and you would just have each of these on its own footprint so that there's no risk of tangling and that kind of thing. I could imagine this could become a disaster; I've done kiting and the number of tangles I've gotten into has been tremendous. So how do you surmount that? Is it that you have to have an area that you exclude other drones from so they don't fly in to each other.
Edgar - There's always a risk. Now it's about how big or how small that risk is. We develop to aeronautical development standards. That means that there is a risk of one in 10 billion or something like that.
Chris - What about noise? Because people are very concerned with the present generation of wind turbines about the noise they make.
Edgar - It is quieter when we're operating in the air it is travelling at about four times the wind speed. When you talk about the turbine blades they go up to 40 times the wind speed, and that's what's making the noise so wind noises we don't expect. We do have a noise at takeoff and landing. We talk about motors, electrical motors to put for a megawatt system 2 tons of weight into the air, but you only have it when it goes up and when it lands and that's maybe once a day, once a week, who tells? If the wind is enough it just stays up there without noise.
46:32 - Returning sensory information
Returning sensory information
with Laura Bücheler, Ghost
A person with nerve damage in their hand, for instance, may not be able to feel if the glass they’re holding is about to slip from their grasp. But Ghost are a company hoping to give back some of that information. They have pressure sensors in a glove which sends signals to vibration motors further up the wearer’s arm or elsewhere on the body, relaying information about what’s going on in their damaged hand. And that’s just the beginning of their work. Adam Murphy spoke to Laura Bücheler, co-founder of Ghost...
Laura - We use any type of complex information, translate them into our haptic language - haptic just means that you can feel it, so instead of listening to something or seeing something we make you feel something, that in turn will activate a few vibration motors placed in a haptic vest that will be vibrating on your back and your brain will learn how to interpret those signals and learn to identify them as the information coming in. So, for example with a prosthesis, people can feel pressure or temperature that is coming into the prosthesis and they will be able to interpret those.
Adam - What exactly have you created with GHOST, and what concrete prototypes do you have?
Laura - We, at the moment, have two prototypes that we showcase. One showing the principle of a glove connecting pressure sensors through vibration motors located on your lower arm, and the other one is the haptic vest that we're actually going to make as a serial product with vibration motors in it to just show how information could be translated into different patterns.
Adam - Say I have a prosthesis and I have this glove on, how is it going to help me? What's it going to do to make my life better I suppose?
Laura - Usually wearing a prosthesis, it's a one-way street. You either have one that just looks like a hand or you have one that you can move, but it does not give you any information, any feedback whether the bottle that you're holding will slip out of your prosthesis or whether you're crushing it.
Adam - What stage are you at with this technology?
Laura - We have a working prototype that shows the principle of it. We are currently working on making that into a smart textile, so no cables - all included in the textile. And working on the haptic language and try to also work on an algorithm that correctly translates the information that comes in into our haptic language.
Adam - What challenges have you found in trying to create this?
Laura - It's actually quite different from what you would expect to what you can feel, you know. How far do you have to place the vibration motors apart from each other, how strong they can vibrate, what can actually be perceived by the human body. And we are mainly also focusing on how to make it most intuitive, to shorten the time that your brain actually has to learn how to interpret the signals and actually that's quite a difficult task.
Adam - How do you go about doing that actually; interpreting buzz means you're holding something?
Laura - We try different vibration patterns and see how well people can pick it up and how fast it takes them to pick it up. Let's say we tried A, B and C and then we realised oh, people learnt A the fastest so let's try to add that into our language and leave B and C to something - it's more complex words so to say.
Adam - And what are you hoping to do in the future? Where are you hoping to take GHOST?
Laura - So we're hoping to take GHOST also to different application fields. So we're currently looking into different ways where we could apply the technology because it's not limited to prosthetics. So, for example, it's not a big jump to think that it's applied from a prosthesis to a robotic arm for example. We could make people who remotely control a robot, feel whatever that robot is doing, or we could even go as far as creating new sensors like, for example, making people feel radiation.
Adam - Oh, so if radiation is coming in even, like could that be nuclear or just solar radiation?
Laura - Whatever would be necessary, it could be both, depending on the sensors that we used to pick it up.
Adam - Could you do that for any kind of sensor? Could you create buzzers for infrared or for UV?
Laura - Exactly!
50:48 - A universal test for cancer
A universal test for cancer
with Karl Bergman & Francesco Gatto, Elypta
One person in three will, at some point in their lives, have a brush with cancer. But very often cancers present at a late stage when they’re much harder to treat. If we could pick them up sooner, or spot when they return after treatment much earlier, the prognosis would be dramatically better. And that’s Elypta are trying to do, they’ve discovered a unique fingerprint of chemical changes present in body fluids when a person has any kind of cancer, providing the possibility of an early warning system for the disease. Chris Smith spoke to Francesco Gatto and Karl Bergman from the Elypta team...
Francesco - It's about discovering the right biomarkers, essentially molecules that we can find in the blood, in the urine that indicate cancer might be present. And we identified a process in metabolism which is essentially how cells nourish themselves. And we know that cancer does it differently.
Chris - When you say you discovered a problem what, so, because of the physical presence of the cancer it churns out molecules or a different combination of molecules that are a different combination to normal. That's what you're saying you're detecting?
Francesco - That is correct.
Chris - And what molecules are you going after?
Francesco - It's complex sugars. They participate in the way the tumour grows in a certain tissue and we detect that kind of process directly in the blood or in the urine.
Chris - And what, because the cells are behaving abnormally, because they're cancerous they make these things which shouldn't normally be there? So that's a hallmark that there must be cancer somewhere?
Franceso - Absolutely. The pattern is actually a bit more complex. That's why we have also software we're developing because it’s not about only creating something new because you have a cancer in place. It's also that cancer needs to modify what's normally available in the body and make it amenable for its own growth.
Chris - And how are you applying this Karl? What's actually happening when you when you test things?
Karl - What we've done is that we've tested this in a wide range of cancers to understand what the potential is and what we find is that this is something that we can use across the cancer spectrum. So what we're doing now is that we're developing assay kits, basically the reagents that we need in order to measure these metabolites in the blood, to commercialise this as a diagnostic test.
Chris - So you would take a blood sample or say a urine sample, you can spot the presence of these abnormal molecules there, that tells you this person might have cancer somewhere but it doesn't tell you where though does it?
Karl - No that's true. It doesn't tell us conclusively where but that's why our first use of this is actually to help patients that have already been diagnosed with a certain cancer. So they've had surgery to remove the cancer and are more or less cured but then they need to come in for a follow-up for many years, and we know what cancer it is if it should come back. So it becomes less of a problem.
Chris - Is there any way, Francesco, to track down the source of these abnormal molecules in the body, so that having just detected them you could then also say well where are they coming from, because that will pinpoint where the tumors might be?
Francesco - The molecules that we're actually tracking and do not directly give information about the tissue where the cancer has started but we are discovering a lot because we're a bit at the forefront of this field. Cancer that come from different types might have slight changes between themselves and we can use this software to capture if you want subtle differences between cancer types, exactly for this purpose.
Chris - One of the crucial things about a screening test, Karl, is that you must not miss any cases. So how sensitive is this test, so that you can say to somebody if they have a negative test this rules out cancer in you?
Karl - So we've done studies in blood samples that we've collected from clinicians and identified that we have a sensitivity across cancer of around ninety-nine percent. So it's highly sensitive to detect early stage cancer.
Chris - So what's the next step then?
Francesco - Well we just received a European Union grant and that will allow us for the first time to roll out these in a very large patient population, across eight to eleven hospitals in the United States and Europe and primarily in kidney cancer, to prove that we can use our test to detect the coming back of cancer after surgery as early as possible, which of course has tremendous impact in the life of these people
Chris - And beyond kidney cancer, because you're looking for something which is a marker more generically of cancer rather than just one specific type of cancer, are you planning to then say, well let's start screening people for lots of different types of cancers? The reason I'm asking the question, Francesco, is that it may well be that some of these kidney cancer patients develop another kind of tumour in the meantime and you might get a signal and it's not because their kidney cancer's come back, it's because they've got lung cancer as well.
Francesco - Right. That's quite rare in the clinic but it may happen. And of course we will most likely detect it as an effect of the fact that our markers are precisely designed for that reason. Now from where we are today to the moment to which we can do screening, there will be a lot of research to be done. But of course Elypta has a quite extensive clinical program. We have already collaborations in prostate cancer and bladder cancer for example. And we did complete a very recent study in lung cancer that has indicated very promising results.
Chris - And is this going to end up being a countertop test in a general practice, as in you could as part of a well man or well woman middle aged health screen you could go and have one of these tests?
Karl - In the first setting, from the patient's perspective you won't notice the difference. You leave your blood but then the blood will be sent to a central laboratory where we collect all the blood samples and do the testing. And the reason for this is it’s quite complicated to do this analysis, it’s a novel method. But as time moves on we are likely to automate this and then bring it so that it becomes available in any clinic.