TINY SENSOR, HUGE FUTURE
Kevin Borras meets Heba Bevan, founder and CEO of revolutionary low-power smart sensor system UtterBerry
While doing my research for this article I read an interview that UtterBerry founder and CEO Heba Bevan did with New Civil Engineer magazine earlier this year. The interviewer remarked on the beautiful, clearly hand-painted tea-cup that Heba drank from during their conversation, relating it to her love of craftsmanship which is reflected in the level of quality engineering that has been invested in her award-winning lightweight miniature low-power intelligent wireless sensor system. Room for another adjective? Thumb-sized, although that clearly depends on the size of your thumb. It’s been described as “fitting into the palm of your hand” which is not inaccurate but I have quite small hands and two Utterberry sensors can fit into the palm of mine and there’d still be room for at least a couple of mints.
I am now looking at that very same tea-cup, seemingly spawned from a similarly delicately crafted Middle Eastern-inspired tea pot that takes centre stage on her expansive desk, together with two laptops housed in cases that look uncannily like ancient scriptures. When I interview someone in their own environment I often get a very strong sense of the type of person I’m dealing with in the first 30 seconds. The word that comes to mind in the presence of the brilliant, young entrepreneur is “details.”
So, how did a young mother of one from Yorkshire come to create one of the most exciting young technology companies in Europe?
“Utterberry is a spin-off from Cambridge University where I was doing my PhD. I realised that the sensors that are being used today in the construction industry were large, consumed a lot of power and had a very minimal capability of performing any complex algorithms or machine learning or anything other than reading data,” she replies, sipping her aromatic tea.
“They are, more than anything, collectors. They collect data and you can pull the data out after a week or sometimes a month.”
At Cambridge Heba studied electronics and computer engineering, receiving a first for her final year project on robotics visualisation. After graduating she worked as a central processing unit engineer for microchip design company, ARM. Returning to academia as a research programmer at Rice University in Houston, Texas, Heba worked on the PACE Project, a major programme for the US defence agency DARPA. Heba completed her PhD at Jesus College, Cambridge, researching low power wireless sensor networks for subterranean mass-rapid transportation applications.
“I spoke with my supervisor and told him that I could develop a sensor that was smaller than a credit card and could harvest the data from different measurements and the sensing capability could be applied to AI or machine learning algorithms, to process the data instantly and provide informative data. The processing is done at the board, at the heart of the sensor. The sensor is low power and harvests different sensing capabilities and compiles and presents informative, useful data. I founded the company in 2013 and I started working on some major projects. One of them was the Post Office Tunnel in, or under, London, one of the oldest tunnels in the UK. It was previously used to carry post across London. Then there was the CE360, which Utterberry won the best innovation award for our work with Crossrail. We monitored what is basically an enormous hole in the centre of London that is completely sealed with no entrance points. Our sensors provided the capability to know exactly what was going on underground from above ground.”
The UtterBerry has been certified as Highly Commended in three Institution for Engineering and Technology (IET) Innovation Awards – the Asset Management Award, Built Environmental Award and Measurement in Action Award – but one of the most tangible benefits it brings about is that an UtterBerry sensor system can be installed in one day by a single person. And, to add to the intrigue, it has vibration at its core.
“The sensor detects minimal physical changes, such as vibration, displacement, pressure, temperature, humidity; it can detect heat in different forms so it can differentiate between a human and a small animal. The data is then identified and filtered to enable you to distinguish different elements such as the time of day, the season, and so on,” Heba explains. “Basically it’s a computer without a screen and a keyboard. People talk about the importance of big data, but what you actually need is informative data. Big data is all very well but often you are presented with so much you don’t know what to do with it.”
Without bringing the conversation down to my level too early, it seems to me that these sensors, when conveyed in a network, simply talk to each other and relay only the most important information to the next sensor in the chain. Unless I’ve really oversimplified things.
“No, not at all,” she replies, somewhat reassuringly. “Our sensors talk to each other, they co-exist as a mesh network or a star network. The sensors work out the best-case scenario for using the least amount of power, hopping between each other, sharing information, and acting as a cluster of families. Let’s say we have a cluster of sensors on Tower Bridge, they will communicate to Blackfriars Bridge and then those sensors can talk to Vauxhall Bridge. While they are talking to each other the clusters share certain information to ensure their algorithms and AI capabilities are working perfectly. In order to communicate with each other our sensors use radio frequency technology. When you have a cluster, like we do in London and in Hong Kong, as another example, through the gateways and the cloud you can have another layer of communication and they can compare each other’s data, for example in terms of movement or vibration.”
Talking of bridges, crossings are one such type of structure where UtterBerry’s ultra-lightweight sensor (it weighs less than 15g) really comes into its own. What might have required the installation of a dozen different sensors as little as five years ago, now requires one human thumb-sized unit to perform all of those tasks and another half-dozen more.
“With bridges, for example a bridge that opens to allow ships through, we can monitor the load on the bridge, we can monitor footfall, the number of cars, the number of bicycles, if the piers are strong enough during the tidal changes, all of those things, in one sensor. All these elements are what makes a bridge ‘come alive’ – if you are monitoring several bridges you can identify exactly when the tide will hit each one. We can also monitor what effects the weight of traffic will have at that time on a certain bridge.”
In terms of disaster relief and emergency evacuation scenarios, such as were executed during Hurricane Katrina in 2005 where hundred of thousands of people were being “strongly advised” to use bridges to leave the Greater New Orleans area, how would an Utterberry sensor network perform?
“There’s two main issues here,” Heba responds without hesitation. “Often when a city has been hit by some kind of disaster, such as an earthquake, the electricity gets cut and the network goes down. I was in Houston, Texas in 2008 during Hurricane Ike and due to the networks being down I lost touch with my family in London for four days. so I’ve been in exactly that situation that you mention. Now, although our sensors are battery powered or can harvest and store energy without being connected to a power supply, in a catastrophe like a hurricane or earthquake you can send data to a mobile phone that’s being used as a base-station, but if the mobile networks are down you have another hurdle to overcome. We don’t transmit on a mobile frequency so it could be used in a different way, as long as we have one mast and one provider still working. However, if the entire city’s communication networks are down it makes it a great deal harder.”
IoT AND THE SMART CITY
It’s not just roads, tunnels and bridges where UtterBerry have cornered the market in an extraordinarily short amount of time. They are also making notable inroads into the fields of smart construction, smart hospitals, smart agriculture, smart trains, smart industry, smart buildings, smart devices and crowd-sensing (also worthy of the smart prefix). Pieced together, the Utterberry Smart City takes on a whole new perspective, a city that isn’t just smart, it knows it’s smart. But, referring back to the company’s interest in smart hospitals, is the age of a large number of the UK’s hospitals a barrier? With some dating back to the mid-19th century, how easy is it to make a building (or in many cases, a bunch of disparate, random structures) smart? Is it not a far simpler task to integrated 21st century technology into 21st century buildings? Retrofit versus embed, if you will.
“In an ideal world the sensor would be integrated into a new build, yes, but we’ve been monitoring a number of Grade I and II listed buildings, such as Tower Bridge, Greenwich Pumping Station, and we can build the sensor in a colour that exactly matches the building and in terms of attaching them to the structures we have done that in a number of tunnels, buildings and bridges in the UK. It’s just a matter of gluing them to the structure,” says Heba, almost dismissively. “A cluster of sensors talking to each other forms a part of the infrastructure to give you the informative data that is needed. To create a network within that existing building is actually super-easy. We can locate the sensors in places that are aesthetically pleasing for visitors, patients and hospital staff. Our sensors can provide way more information than they would be required to produce in a construction environment. For example, we can measure the amount of people that have used certain corridors to monitor infection, we can sense seat occupancy, whether fire doors are closed or open – there are many different things we can achieve desired outcomes in a healthcare setting.”
A few years ago I was discussing crowd sensing and sourcing with the University of Leeds’ Susan Grant-Müller and in particular a project they were working on in Madrid where social media posts that contained references to transport were monitored after Atletico Madrid matches at the Vicente Calderon Stadium to ascertain potential traffic bottlenecks, public transport overcrowding and so on. With UtterBerry listing crowd sensing as one of its core competencies, I was curious to know if this was something that Heba Bevan was aware of.
“Firstly yes, it is and secondly I am glad you mentioned Leeds favourably as that’s where I was born!” she replies, wistfully. “But in terms of people counting we worked on a lengthy project with Network Rail and the Rail Safety & Standards Board measuring the number of people who were using level crossings, detecting the percentage of children and what types of animals were using the level crossings. The technology being used typically is very expensive and not accurate enough. We collect data in an anonymous way – we don’t take pictures or capture sound. We’ve also used this with CrossRail as part of an Innovate18 project. This was led by CrossRail’s CEO, Andrew Wolstenholme, and the project focuses on anonymously counting the number of people in tunnels so the exact numbers and locations can be measured when there’s a fire, for example. We came up with a solution that was to utilise our existing sensors that have been monitoring structural behaviour and changes in the environment so that they also count the number of people in particular zones, giving those people information about which is the optimum exit for them to use. The nearest exit may not necessarily be the safest one in certain situations.
“We take a huge amount of measurements, such as vibration, signal propagation, heatwave, structural movement, count the number of people in the tunnel and in which direction they are moving and provide those people with incredibly accurate and instantly useful information. We’re using that methodology at two level crossings in West Sussex.”
Heba Bevan’s career path, it would seem, was mapped out from quite an early age. Her parents are academics, and as a child she lived in several countries as her parents’ careers took them around Europe. While at university she worked as an intern at Cambridge-based ARM, a British multinational semiconductor and software design company, owned by SoftBank Group and its Vision Fund. By the time she graduated, being offered a full time job as an integral designer of the ARM Cortex processors was the natural next step.
At just 22, Bevan was working for ARM in Silicon Valley and “going to meetings with clients such as Apple and Google on my own.” Next was a return to academia as a research programmer at Rice University in Houston, Texas, working on DARPA’s PACE Project. Now married, Heba and her husband moved back to the UK where she decided to do a PhD at Jesus College, Cambridge, to research low power wireless sensor networks for subterranean mass-rapid transportation applications. “The reason I chose Cambridge,” she insists, clearly still pleased with her choice, “was that at most universities, you don’t own your intellectual property, but at Cambridge you do. And this is where the concept of UtterBerry came to life.”
Achieving so much so soon can, in some instances, see the company rest on its laurels and while it’s sitting back in its metaphorical deckchair drinking a metaphorical cocktail, it finds it’s so metaphorically relaxed that it’s in no position to do anything other than watch as its competition overtakes it and disappears into the distance. There is absolutely zero chance of that happening to UtterBerry. Zero.
“This is only the beginning,” she states. “The ambition is to have a very strong technology base in the UK. To be honest the UK is lacking in technology companies. Often when they get to the stage that we are at they get sold. We’re very proud that we are a UK company, we manufacture everything in the UK and all of our technology showcases have been done here too, although we are now working with Mass Transit Railway MTR in Hong Kong and also in China and Australia. We’re expanding and looking to have a global presence. That’s not going to be easy as we want to work in a number of different sectors outside the UK. We’re not done. This is not done. You can’t just be satisfied with what you’ve done. We’ve come this far but now we want to go even further. I’ve said this many times but it is not just about the idea, it is about the person leading the idea and the person who gets their message across the entire spectrum of the industry, by taking their vision and making it everybody’s.”
Having been the only female on her degree course (“There was only one the year after and that was my sister”) Heba is understandably and palpably passionate about introducing technology at an early age, with extra focus on engaging with girls, and has developed a school’s package to allow school children to develop software using her range of sensors. A software integration package, allowing third party developers to develop apps using the UtterBerry sensor range is a medium-term project.
“I really want to get more women at least interested in engineering, starting from primary school. I’m a STEM (Science, technology, engineering, and mathematics) ambassador in primary and secondary schools and there is so much talent and potential in this country and the UK has so much to offer the world.”
THEM AND U.S.
Having worked on both sides of the Atlantic, Heba remains unconvinced that the UK’s attitude to start-ups is entirely correct.
“Silicon Valley will fund interesting, cool or innovative tech, with utility and purpose being secondary. There is also no requirement for profit; in fact companies with forecasts that show no profit are often more valuable than those that do,” she says sagely. “The UK has a much more conservative approach to investment. The utility of the technology needs to have a strong business case, demonstrable demand and a clear profit forecast. Furthermore, a UK tech start-up needs to be successful straight away, with failure considered a major negative. Failure in Silicon Valley is celebrated, with pivoting seen as a badge of honour. ‘Fail fast, fail often is a favourite of the Valley mantras’.”
One of the starkest differences between the two global centres of technical excellence is in the area of collaboration. A Silicon Valley start-up does not need to reinvent the wheel when their technology has ‘gaps’, instead they will find a partner to work with, or a software service product to license/purchase. A great example of an ‘open’ start-up is Twitter, where they built the shared platform first and then developed the user-interface over the top. Does the British approach hamper progress?
“What I will say is that the UK tech scene is much more independent, with companies developing end-to-end products without considering third party accessibility.”
PREPARE FOR IMPACT
Heba believes that Smart Infrastructure is the fundamental base from which a smart city can be built.
“Without smart buildings, smart roads, smart tunnels, we cannot hope to be a truly connected, smart city. A key component of a fully connected city is data, and the insights that can be gathered from it.”
The world’s smallest sensor is having a big impact on our world – in fact you could say that UtterBerry has a clearly fruitful future ahead of it.
Additional reporting by Darren Price.
Kevin Borras is editor-in-chief of Thinking Highways and Thinking Cities
Darren Price is technology editor of Thinking Highways and Thinking Cities
Find out more about UtterBerry at http://www.utterberry.com