Sensor fusion refers to the capability to combine input from different radars, LiDARs, cameras to create a single image of an environment around a vehicle.Such a process is required as sensor imperfections are a certainty. No system can afford the luxury of operating under pristine, lab-grade conditions, where distractions never happen.Aside from external disturbances in working conditions, there are also the sensor’s own issues—drifting, inaccuracies, or just plain crazy black box errors.Sensor fusion is one component you’ll use to design this ‘better’ system. <div class="fr-img-space-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE3MzY1OTQzNzk4LWV5SmlkV05yWlhRaU9pSjNaWFp2YkhabGNpMXdjbTlxWldOMExXbHRZV2RsY3lJc0ltdGxlU0k2SW1aeWIyRnNZUzh4TlRnek1qUXpPRGd3TURRd0xVRjFkRzl1YjIxdmRYTWdWbVZvYVdOc1pTQlRaVzV6YjNKeklEQXhJREV3TWpSd2VDNXFjR2NpTENKbFpHbDBjeUk2ZXlKeVpYTnBlbVVpT25zaWQybGtkR2dpT2prMU1Dd2labWwwSWpvaVkyOTJaWElpZlgxOS53ZWJwIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" style="width: 788px;" class="fr-fic fr-dib" alt="Sensors used in autonomous vehicles">An example of typical sensors used to perceive the environment. Note that various vehicle manufacturers may use different combinations of sensors and might use all of the displayed sensors. For example, increasingly multiple smaller LIDAR sensors are being used, and long-range backward-facing RADAR can be incorporated to cover situations like highway lane changing and merging. The placing of the sensors can vary as well. Image: Wevolver<h3 class="fr-img-space-wrap2">What is sensor fusion?</h3></div>You might be more familiar with sensor fusion if I call it “data fusion,” “information fusion,” or even “multi-sensor integration.”These phrases are often mentioned like they’re the same thing.Spoiler alert: they’re not.But a sensor fusion system can involve all these principles at the same time.By definition, sensor fusion is a process to join the measurement data from multiple measurement sessions, be it from different sensors or different instances, to get better measurement data than if you’re using a single sensor as your source of truth.‘Better’ in this sense can be something as simple as a more accurate representation, more complete data, or even inferred information that isn’t directly measured by current sensors.<h3 dir="ltr">4 reasons to use sensor fusion</h3>Let’s face it:Issues come up when you use sensors.Whether it’s something caused by imperfection of the sensor itself, like noise and drifting, or an external factor, like an extraordinary operating condition.Certain as they are, you can side-step these issues by strengthening your data stream with multiple sensors, whether they’re duplicates or heterogeneous sources.Let’s say you want to measure the vibration of a bridge when a vehicle passes.You left an accelerometer overnight, which transmits the data live to a computer in your office.When you saw the data the next morning, you noticed that there’s an outlier in your data—a sudden burst in your accelerometer’s wavelength.What happened there?Did a particularly big truck just happen to pass by? Is it just plain sensor craziness? Or is there a cyclist who knocked the accelerometer out of its position?Unfortunately, since you’re only using one sensor in this case, you’ll never know. And you’ll never know if the data your sensor recorded is accurate either.This is where using multiple sensors and a sensor fusion algorithm helps.<ol><li dir="ltr">Adding a second (and third) sensor lets you compare the data. This improves the accuracy of your measurement results, so you can hypothesize about what happened at the bridge that night.</li><li dir="ltr">If there are outside interferences disturbing your sensor’s performance, you can use a second sensor that isn’t affected by this disturbance. Thus, maintaining the consistency of your data stream.</li><li dir="ltr">If by any chance, your sensor is experiencing a breakdown in the middle of the operation, you don’t have to suffer from data loss as there are backup sensors available.</li><li dir="ltr">Get additional insights using the same set of sensors, reducing the complexity, number of components, and cost of operation for your system.</li></ol>The margin of error for your system gets smaller when you use sensor fusion to merge sensory data from multiple sources, as opposed to trusting a single sensor to perform exceptionally well throughout the entire operation.<h2 dir="ltr">Types of sensor fusion systems</h2>When observed from a high-level, the setup you’ll need for a sensor fusion network is fairly simple.You have the sensors, which give an input of raw data to the estimation framework, where the algorithm will transform your raw data into a specific output.Depending on the algorithm you’re using, the output could be an entirely unique set of information, or it could be a more accurate version of your raw data.However, the algorithm, the architecture of the overall system, and the sensor setup will depend on what sensors you have and what information you want to get out of this sensor fusion system.Let’s start with the easier part and get to know the types of sensor fusion systems, based on the setup, architecture type, and levels of fusion.<h3 dir="ltr">Sensor setup</h3>Arguably the simplest categorization, there are three commonly used sensor fusion configurations: competitive (or redundant), cooperative, and complementary. <a href="https://link.springer.com/chapter/10.1007/978-1-4613-8997-2_7" target="_blank">(Durrant-White, 1990)</a><ol><li dir="ltr">Competitive In a competitive or redundant setup, your sensors observe the same object. By having multiple sensors set up to observe the same state, you can improve the reliability and accuracy of your system, and the confidence level of your measurement data.</li><li dir="ltr">Cooperative Just like a competitive setup, cooperative systems also observe the same object. Rather than using the data to decide which data stream should be passed on, you can use a cooperative setup to get new insights into your observed environment by combining the raw data.</li><li dir="ltr">Complementary A complimentary setup observes two different objects. Similar to a cooperative setup, merging the data lets you form a global point of view of the observed environment.</li></ol><h3 dir="ltr">Architecture type</h3>Based on its architecture, there are three main types of sensor fusion systems: centralized, decentralized, and distributed. <a href="https://www.hindawi.com/journals/tswj/2013/704504/" target="_blank">(Castanedo, 2013)</a>There is also a hierarchical type, which combines the decentralized and distributed system through a hierarchy, but let’s stick to the three main types for this brief explanation.<ol><li dir="ltr">Centralized Within a centralized architecture, the data fusion process is done directly on the control system. This simplifies the architecture, but puts a strain on the control system, making it slower and less efficient.</li><li dir="ltr">Decentralized In a decentralized architecture, your measurement data is combined at a data fusion node, with feedback from your local network of sensors and other data fusion nodes in your system. While this is more efficient than a centralized system, your communication line gets complicated quickly if you use a lot of sensors and data fusion nodes in your system.</li><li dir="ltr">Distributed A distributed system processes your data locally at the source before sending it into an intermediate fusion node, which then produces a global view of your observed environment.</li></ol><h3 dir="ltr">Levels of data fusion</h3>Depending on your application, your sensor measurement data will transform into different levels of data as it goes through your system. Ackoff states that the <a href="http://www.systems-thinking.org/dikw/dikw.htm" target="_blank">hierarchy of your data</a> can take the form of data, information, knowledge, and wisdom. <div class="fr-img-space-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE2NzUxMjM4NDc3LURJS1cgcHlyYW1pZC5wbmciLCJlZGl0cyI6eyJyZXNpemUiOnsid2lkdGgiOjk1MCwiZml0IjoiY292ZXIifX19" style="width: 690px;" class="fr-fic fr-dib" alt="DIKW hierarchy wisdom hierarchy knowledge hierarchy information hierarchy">DIKW Pyramid&nbsp;In this pyramid, data is raw data. Or basically just an input from your measurement system —whether it’s numbers, symbols, signals, or others—that you can’t really make sense of just yet.</div>The next level, information, is still data but this time tagged with the relevant context so it becomes more useful and easily visualized.While knowledge is obtained by assembling the pieces of information and relating it to what you want to find out. In short, knowledge is where the data starts to become useful for you.The top of our pyramid, wisdom, is a little harder to define. But just like everything else before it, it builds on knowledge. This time to ask more abstract questions like ‘why’ and eventually get a decision.This DIKW ties in neatly to the three main levels of data fusion. Based on the levels of data fusion, there are three categories of sensor fusion systems:<ol><li>Low-level fusion moves at a data level. It takes your raw data as input and returns a more accurate version of your raw data as output.&nbsp;</li><li>Mid-level fusion moves at the feature level. It takes observed features as input and produces a new feature as output. Here, features are just information inferred from a collection of raw data, including contextual information processed from various sensors or a priori information.</li><li>High-level fusion combines decisions, represented from sources as input, to produce a more accurate decision as output.</li></ol>As you can see, this categorization only moves from one level to the next. The problem is when you have a system that crosses levels. Here, you can use <a href="https://www.sciencedirect.com/science/article/pii/B9780128111536000014" target="_blank">Dasarathy’s categorization</a>, which uses input and output data format as the parameter. <div class="fr-img-space-wrap"><img src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjE2NzUxNzIxNjYwLVNlbnNvcl9GdXNpb25fQ2xhc3NpZmljYXRpb25fYnlfRGFzYXJhdGh5LnBuZyIsImVkaXRzIjp7InJlc2l6ZSI6eyJ3aWR0aCI6OTUwLCJmaXQiOiJjb3ZlciJ9fX0=" style="width: 650px;" class="fr-fic fr-dib" alt="sensor fusion classification by Dasarathy">Classification by Dasarathy<h2 class="fr-img-space-wrap2">&nbsp;</h2><h2 class="fr-img-space-wrap2">How sensor fusion prevents cars from crashing</h2></div>Autonomous vehicles need a lot of input to get a sense of their ever-changing environment.To prevent crashes, the system needs a model of the external and internal state, which will need to be measured using a combination of several sensors, including radar, lidar, and cameras.Naturally, each sensor has its own strength and weakness. To get an accurate depiction of the environment and improve the robustness of the system, you’ll need to combine the data from each sensor.For instance, you need cameras to get more context about the environment and classify objects. On the other hand, lidar and radar are better at determining the distance of the car from an object.The table below depicts how these three sensors can benefit from each other. <table style="width: 100%;"><tbody><tr><td style="width: 19.8985%;"> </td><td style="width: 20%; text-align: center;">RADAR</td><td style="width: 20%; text-align: center;">LiDAR</td><td style="width: 20%; text-align: center;">CAMERA</td><td style="width: 20%; text-align: center;">FUSION</td></tr><tr><td style="width: 19.8985%;">Object detection</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Pedestrian detection</td><td style="width: 20%; vertical-align: middle; text-align: center;">-</td><td style="width: 20%; vertical-align: middle; text-align: center;">O</td><td style="width: 20%; vertical-align: middle; text-align: center;">+</td><td style="width: 20%; vertical-align: middle; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Weather conditions</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">-</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Lighting conditions</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">-</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Dirt</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">-</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Velocity</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Distance - accuracy</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Distance - range</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Data density</td><td style="width: 20%; text-align: center;">-</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Classification</td><td style="width: 20%; text-align: center;">-</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td style="width: 19.8985%;">Packaging</td><td style="width: 20%; text-align: center;">+</td><td style="width: 20%; text-align: center;">-</td><td style="width: 20%; text-align: center;">O</td><td style="width: 20%; text-align: center;">+</td></tr><tr><td colspan="5" style="width: 99.8731%;">+ = Strength; O = Capability; - = Weakness</td></tr></tbody></table> There are many more applications of sensor fusion in autonomous vehicles. It’s also needed for many other safety features required by an autonomous vehicle, such as 3D object detection, tracking, and occupancy mapping. <a href="http://users.isy.liu.se/rt/lundquist/Publications/Lundquist2011.pdf" target="_blank">(Lindquist, 2011)</a><h2 dir="ltr">Want to use sensor fusion in your next project?</h2>So now that you’ve peeked into the world of sensor fusion, there’s one more topic we haven’t covered yet. Just because it’s such a broad topic.Your sensor fusion algorithm, the core of your sensor fusion system.It will largely depend on what you’re trying to do. The algorithm you’ll need to get more precise data for your lab-contained robots will most likely differ from one to prevent crashes on autonomous cars.After you’ve decided on the requirement of your system and which sensors to use, you can start thinking about which algorithm you’ll use to create a sensor fusion system that matches your needs.To get a basic idea about what kind of algorithm you need, you can start with the most popular filters for sensor fusion: <a href="https://www.mathworks.com/videos/series/understanding-kalman-filters.html" target="_blank">the Kalman filter</a>, an estimation framework widely used for its predictive capabilities.<h2 dir="ltr">Bibliography</h2>Castanedo, Federico. 2013. “A Review of Data Fusion Techniques.”&nbsp;The Scientific World Journal&nbsp;2013, no. 704504 (October): 19. https://doi.org/10.1155/2013/704504.Durrant-White, H.F. 1990. “Sensor Models and Multisensor Integration.” In&nbsp;Autonomous Robot Vehicles, 73-89. N.p.: Springer. https://doi.org/10.1007/978-1-4613-8997-2_7.Galar, Diego, and Uday Kumar. 2017. “Chapter 1 - Sensors and Data Acquisition.” In eMaintenance: Essential Electronic Tools for Efficiency, 1-72. N.p.: Academic Press. <a href="https://doi.org/10.1016/B978-0-12-811153-6.00001-4" target="_blank">https://doi.org/10.1016/B978-0-12-811153-6.00001-4</a>.Lundquist, Christian. "Sensor fusion for automotive applications." PhD diss., Linköping University Electronic Press, 2011

Introducing sensor fusion, how it works, and why you should use sensor fusion in your next system.

Sensor Fusion: Everything You Need to Know

<header>Munich is home to the world's first fully automated sensor network for measuring urban greenhouse gas emissions based on ground-based remote sensing of the atmosphere. It has been developed by scientists in the group of Jia Chen, Professor of Environmental Sensing and Modeling at the Technical University of Munich (TUM). Now, anyone can view the measurement data via an Internet platform.</header><section>The sensor network MUCCnet (Munich Urban Carbon Column network) consists of five high-precision optical instruments that analyze the sun’s light spectra. They measure the concentration of the gases carbon dioxide (CO2), methane (CH4) and carbon monoxide (CO). Since each gas has its own unique spectral "fingerprint", concentrations of these gases can be determined in the columns of air between the instruments and the sun. &nbsp;"By measuring a vertical column of the atmosphere, local disturbances, such as the disproportionate influence of neighboring stacks, can be removed. Therefore, this type of greenhouse gas balancing is considered particularly robust and accurate," says&nbsp;<a href="https://www.professoren.tum.de/en/chen-jia" target="_blank">Prof. Jia Chen</a>.<h2>Measurements at five locations in and around Munich</h2>One of MUCCnet's measurement devices is located on the main campus of TUM and measures inner-city concentrations. Four other devices are located at the Munich city borders in all four cardinal directions (north: Oberschleissheim, east: Feldkirchen, south: Taufkirchen and west: Graefelfing).Chen explains the principle in simple terms: "We set up one sensor upwind from the city and the second downwind. So any increase in gases between the first sensor and the second must have been generated from inside the city." To cover as many wind directions as possible, there is a sensor in each cardinal direction. With the input of the sensor data and meteorological parameters, high-performance computers can create a spatially resolved emission map of the city.<h2>Using measured data to improve the calculated emission figures</h2>Under the Paris Climate Agreement, atmospheric measurements are not required to meet emissions targets. “Instead, the emissions numbers we hear in the news are based on calculations,” explains Prof. Chen.Among other things, this makes it impossible to detect so far unknown sources - such as leaks in gas pipelines. Therefore, Prof. Jia Chen's team and project leader Florian Dietrich created MUCCnet to measure emissions with high precision, which can reduce inaccuracies in calculations.<h2>Corona lockdown as a natural experiment for the measurement data series</h2>The current Corona crisis provides a useful natural experiment for researchers because as a result of the two German lockdowns in spring 2020 as well as winter 2020/21 and severe air traffic curtailment, there have been changes in urban greenhouse gas emissions, which can be used to validate measurements as well as atmospheric transport models.Unfortunately, the lifetime of CO2 is very long (several hundred years) and measurement results show that even such a drastic global event as this pandemic has not stopped the annual increase of CO2 concentration in the atmosphere.<h2>Measurement data can be accessed online</h2>Since the start of 2021, the researchers have operated a&nbsp;<a href="http://atmosphere.ei.tum.de/" target="_blank">website&nbsp;</a>which not only makes measurement data available to everyone, but also explains the devices used and the principles employed to gain the data. Interested parties can find absolute values of greenhouse gas concentrations on the portal and can, for example, draw comparisons between stations at different locations."Since climate change is a global problem, the Munich network should only be the first step," says Prof. Chen. In the future, Chen's team plans to use measurements from existing greenhouse gas satellites to expand the methods and models developed in Munich worldwide and thus make a decisive contribution to understanding and solving the climate problem.</section>

Prof. Jia Chen and project manager Florian Dietrich at one of the sensors for measuring urban greenhouse gas emissions. Image: A. Heddergott / TUM

Munich is home to the world's first fully automated sensor network for measuring urban greenhouse gas emissions based on ground-based remote sensing of the atmosphere.

Unique sensor network for measuring greenhouse gases

As Ingo Burgert and his team at ETH Zurich and Empa have proven time and again: wood is so much more than just a building material. Their research is geared to enhancing the properties of wood to fit it for entirely new applications. For instance, they have already developed high-strength, water-repellent and magnetisable wood.Now, together with the Empa research group led by Francis Schwarze, the team has used one chemical and one biological process to generate electrical voltage from a type of wood sponge. In doing so, they amplify what is known as the “piezoelectric effect” of wood.<h2>Compression creates voltage&nbsp;&nbsp;</h2>When a piezoelectric material is elastically deformed, it generates an electrical voltage. Measurement technology, in particular, exploits this phenomenon by using sensors that generate a charge signal when mechanically stressed. However, many of the materials often used for these sensors are unsuitable for biomedical applications. Lead zirconate titanate (PZT), for example, cannot be used on the skin due to its toxic lead, and must be specially disposed of.Wood also has a natural piezoelectric effect, but only generates a very low electrical voltage. If one wants to increase the voltage, the chemical composition of the wood must be changed – and this also makes it more compressible.<h2>From block to sponge</h2>In order to convert wood into an easily formable material, one component of the cell walls must be dissolved out. Wood cell walls consist of three basic substances: lignin, hemicellulose and cellulose. “Lignin is the stabilising substance that trees need to grow tall. Without lignin, which connects the cells and prevents the stiff cellulose fibrils from buckling, this would not be possible,” states Burgert.A few months ago, Jianguo Sun, a doctoral student in Burgert’s team, together with colleagues from ETH and Empa, showed in a study published in ACS Nano how wood can be made deformable if the lignin is removed chemically. As a result, its piezoelectric effect is enhanced.The researchers achieved this “delignification” by placing wood in a mixture of hydrogen peroxide and acetic acid. The acid dissolves the lignin, leaving a framework of cellulose layers. “The process retains the hierarchical structure of wood and prevents disassembly of the individual fibres,” Burgert explains.In this way, a piece of balsa wood becomes a white, wooden sponge, made up of layer upon layer of thin cellulose. The sponge can simply be compressed and then returns to its original shape. “The wood sponge generates an electrical voltage 85 times higher than that of native (untreated) wood,” says Sun.<h2>A mini-generator in the wooden floor</h2><div class="fr-img-space-wrap"><img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1615815097159-image.imageformat.fullwidth.1318208763.jpg" style="width: 788px;" class="fr-fic fr-dib">When the stabilising lignin is removed from the balsa wood, flexible cellulose layers remain (centre). Compression then creates an electrical voltage. (Visualisation: ACS Nano / Empa).&nbsp;The team subjected a test cube with a side length of approximately 1.5 cm to around 600 load cycles. The wooden sponge proved surprisingly stable: for each load, the researchers measured a voltage of approximately 0.63 volts, which would be appropriate for a sensor. In further experiments, the team tested the scalability of this mini-generator. If 30 such wooden blocks are connected up and evenly loaded with the body weight of an adult, enough electricity is generated to power a simple LCD display.<h2>Treatment with fungus instead of chemicals</h2>In a follow-up <a href="https://advances.sciencemag.org/content/7/11/eabd9138" target="_blank">study</a> just published in Science Advances, the ETH-Empa research team went one step further, seeking to produce the wooden sponge without using chemicals. The researchers found the solution in nature: the fungus Ganoderma applanatum causes white rot in wood, and degrades the lignin and hemicellulose particularly gently. “Although the electrical voltage generated was lower in initial tests than with chemically treated wood, the fungal process is more environmentally friendly,” says Burgert.There are clear advantages to such a simple, renewable piezoelectric system. The researchers see various potential applications for the wood sponges – for example, as sustainable building materials that harvest energy in the use phase, or skin-friendly pressure sensors for medical purposes.However, there are still several steps to go before piezo wood can be deployed as a biosensor, or even as an electricity-harvesting parquet floor. Burgert and his colleagues are now exploring with various cooperation partners how to adapt the technology for industrial applications.This is a slightly modified version of an article originally published by <a href="https://www.empa.ch/web/s604/strom-aus-dem-parkett" target="_blank">Empa</a>.<hr><h2>Reference</h2>Sun J, Guo H, Schädli GN, Tu K, Schär S, Schwarze F, Panzarasa G, Ribera J, Burgert I. Enhanced mechanical energy conversion with selectively decayed wood. Science Advances (2021); doi: <a href="http://dx.doi.org/10.1126/sciadv.abd9138" target="_blank">10.1126/sciadv.abd9138</a> &nbsp; Sun J, Guo H, Ribera J, Wu C, Tu K, Binelli M, Panzarasa G, Schwarze F, Wang Z, Burgert I. <a href="https://pubs.acs.org/doi/10.1021/acsnano.0c05493" target="_blank">Sustainable and biodegradable wood sponge piezoelectric nanogenerator for sensing and energy harvesting applications</a>. ACS Nano 14, 14665–16474 (2020).</div>

Even a little pressure can generate usable energy in the wooden sponge. (Photograph: ACS Nano / Empa)

Researchers at ETH Zurich and Empa have chemically modified wood and made it more compressible, turning it into a mini-​generator. When compressed, it generates an electrical voltage. Such wood could serve as a biosensor or as a building material that harvests energy.

Voltage from wood

In this episode, we talk about a novel approach to detect Alzheimer's using retinal images, monitoring stress levels using sweat, and a new method of processing biomass which leads to better bioplastics. As always, you can find these and other interesting &amp; impactful engineering articles on Wevolver.com.&nbsp;<iframe height="200px" width="100%" frameborder="no" scrolling="no" seamless="" src="https://player.simplecast.com/72b0b057-b0ea-4dff-bc28-74791ef72352?dark=false" class="fr-draggable" data-lf-form-tracking-inspected-kn9eq4roawkarlvp="true" data-lf-yt-playback-inspected-kn9eq4roawkarlvp="true" data-lf-vimeo-playback-inspected-kn9eq4roawkarlvp="true"></iframe>EPISODE NOTES &nbsp;(0:45) Diagnosing Alzheimer’s At A Glance -&nbsp;Researchers at Duke University have developed a machine learning model capable of successfully differentiating between healthy patients and those with Alzheimer’s using retinal images.(7:00) The Stress You Sweat -EPFL and Xsensio have collaborated to create a wireless sensor using graphene capable of monitoring cortisol levels secreted via sweat in quasi real-time.&nbsp;(12:30) Bloom’s Better Bioplastics -Bloom Biorenewables biomass processing method increases the efficiency of resource extraction from 40% to 75% and enables production of high performance and biodegradable bioplastics.--About the podcast:Every day, some of the most innovative universities, companies, and individual technology developers share their knowledge on Wevolver. To ensure we can also provide this knowledge for the growing group of podcast listeners, we started a collaboration with two young engineers, Daniel Scott Mitchell &amp; Farbod Moghaddam who discuss the most interesting content in this podcast series.&nbsp;To learn more about this show, please visit the&nbsp;<a href="https://www.wevolver.com/profile/the.next.byte" target="_blank">shows page</a>. By following the page, you will get automatic updates by email when a new show is published.Be sure to give us a follow and review on <a href="https://podcasts.apple.com/nl/podcast/the-next-byte/id1548255861?l=en" rel="nofollow" target="_blank">Apple</a> podcasts, <a href="https://open.spotify.com/show/6lPPsRtegnivsRUEsQ09R7?si=IPMiah7xT_apJtPjqvXMlA" rel="nofollow" target="_blank">Spotify</a>, and most of your favorite podcast platforms!

In this episode, we talk about a novel approach to detect Alzheimer's using retinal images, monitoring stress levels using sweat, and a new method of processing biomass which leads to better bioplastics.

Podcast: Diagnosing Alzheimer's At A Glance, The Stress You Sweat, Bloom's Better Bioplastics

Last year's release of <a href="https://www.xsens.com/news/xsens-mvn-analyze-2020-advances-sports-science-and-research-for-human-motion" rel="noopener" target="_blank">Xsens MVN 2020</a> updated our world-leading inertial motion capture system with a host of game-changing new features, such as remote on-body recording (rOBR), object tracking, 3D positional aiding, and much more. Already, <a href="https://www.xsens.com/products/mvn-analyze" rel="noopener" target="_blank">MVN Analyze</a> users have begun harnessing these new features to achieve incredible results – one such user is Dr. Alex Moorhead, a research scientist and post-doctoral research fellow at the <a href="https://www.polimi.it/en/" rel="noopener" target="_blank">Politecnico di Milano</a>, where he has utilized the rOBR feature to conduct professional sports analysis on junior tennis players.Completing his master’s degree at the University of Colorado while working at the US Olympic Training Center, Dr. Moorhead has since researched the biomechanics of a broad range of professional sports, completing over 1000 tests and analyses for athletes ranging from swimming, wrestling, track and field, pentathlon and many others – his most recent focus being the performance of professional junior tennis players in Rome. We spoke with Dr. Moorhead about his research into these pro tennis players, and how Xsens MVN Analyze has revealed detailed insight into performance enhancement and injury prevention.&nbsp;Precise analysis with Xsens MVN AnalyzeAfter completing his Ph.D. in Integrated Biomedical Research at the University of Milan, Dr. Moorhead began working at Politecnico di Milano where he was introduced to Xsens when they visited to present their technology. As his work became more widely known, he was invited to research a team of Italy’s very best junior tennis athletes – they were seeking expertise and technology that could guide the enhancement of their athletic performance and technique. With knowledge of various research tools used to study biomechanics, Dr. Moorhead recognized this as a great opportunity to utilize Xsens MVN Analyze for the research. The ability to use the Xsens inertial motion capture suits anywhere with precision allows Dr. Moorhead to capture the dynamic athleticism required in tennis whilst his subjects are on the court. &nbsp;“Before I had even spoken with Xsens to order the suits, I knew it would be the right tool for the job,” explained Dr. Moorhead. “One of the things I like about the Xsens suit is it holds the sensors in place, providing accurate motion capture &nbsp;without the need for frequent recalibration, and yet, it allows the athletes to move without feeling restricted. As soon as they had the suits on, they had no issues at all. The MVN Analyze software is also perfect for the post-processing of the data,” he added. <img width="500" alt="_1081035" src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjEyNTMzMzg0MjI2LTE2MTI1MzMzODQyMjYucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" srcset="https://www.xsens.com/hs-fs/hubfs/_1081035.jpg?width=250&amp;name=_1081035.jpg 250w, https://www.xsens.com/hs-fs/hubfs/_1081035.jpg?width=500&amp;name=_1081035.jpg 500w, https://www.xsens.com/hs-fs/hubfs/_1081035.jpg?width=750&amp;name=_1081035.jpg 750w, https://www.xsens.com/hs-fs/hubfs/_1081035.jpg?width=1000&amp;name=_1081035.jpg 1000w, https://www.xsens.com/hs-fs/hubfs/_1081035.jpg?width=1250&amp;name=_1081035.jpg 1250w, https://www.xsens.com/hs-fs/hubfs/_1081035.jpg?width=1500&amp;name=_1081035.jpg 1500w" sizes="(max-width: 500px) 100vw, 500px" class="fr-fic fr-dii" style="width: 778px;">Using the Xsens suits, Dr. Moorhead was able to obtain clean data from the tennis athletes in real tennis court environments and record accurate, repeatable results. Vital techniques, such as the serve, can be analyzed in realistic scenarios.&nbsp;“From the range of movement and agility testing to actual serves, everything is recorded on the court. We’re trying to find the correlation between service speed, accuracy, and the range of motion in the shoulders – both external and internal with the full extension of the arm and the hip flexors. More than 90% of the tennis match points are decided from the serve or return of serve. So if you can get the most powerful and accurate serve possible, you have a huge advantage,” he explained.Remote On-Body Recording<img width="500" alt="_1081028" src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjEyNTMzMzgzODk3LTE2MTI1MzMzODM4OTcucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" srcset="https://www.xsens.com/hs-fs/hubfs/_1081028.jpg?width=250&amp;name=_1081028.jpg 250w, https://www.xsens.com/hs-fs/hubfs/_1081028.jpg?width=500&amp;name=_1081028.jpg 500w, https://www.xsens.com/hs-fs/hubfs/_1081028.jpg?width=750&amp;name=_1081028.jpg 750w, https://www.xsens.com/hs-fs/hubfs/_1081028.jpg?width=1000&amp;name=_1081028.jpg 1000w, https://www.xsens.com/hs-fs/hubfs/_1081028.jpg?width=1250&amp;name=_1081028.jpg 1250w, https://www.xsens.com/hs-fs/hubfs/_1081028.jpg?width=1500&amp;name=_1081028.jpg 1500w" sizes="(max-width: 500px) 100vw, 500px" class="fr-fic fr-dii" style="width: 778px;">One of several new features launched with MVN 2020, <a href="https://www.xsens.com/blog/remote-on-body-recordingrobr" target="_blank">rOBR</a> (Remote On-Body Recording) became an integral part of Dr. Moorhead’s research. The rOBR feature allows full-body motion data to be tracked and recorded using body packs worn by the user, rather than being tethered to a workstation or laptop. This makes it possible to analyze multiple athletes simultaneously and also means data from the body packs can be transferred to the computer later for analysis. The latter made Dr. Moorhead’s research quicker and more efficient. &nbsp;&nbsp;“rOBR is great. You can watch this data as it’s collected live and stop and cut while the athlete is performing. I ended up transitioning to only using rOBR because of how efficient it was. It makes calibration quicker, it makes it easier to interact with the athlete, and it’s quite nice when you have 12 athletes completing multiple movements. The fact that the data is recorded and downloaded later cuts a lot of valuable time” explained Dr. Moorhead.Dr. Moorhead uses Xsens MVN Analyze to process his data, with as many as 100 files processed overnight, all before inputting the data into MATLAB software for analysis.&nbsp;“rOBR will probably be my primary means of using the software in the future. It was so accurate that I didn’t need to check that the movements were fine while assessing. You can interact with the athlete to an even greater extent and you don’t need to monitor a computer the whole time,” he said. &nbsp;Preventing shoulder injury and improving serving techniqueBy studying the mechanics of the movement, he realized that the serving motion utilizes many of the same muscles which are activated during an overhead squat (squatting with a barbell fully extended above the head) while also giving insight into the athlete’s active range of motion. By studying the athlete’s overhead squat with an empty barbell while in the Xsens suit, Dr. Moorhead could assess individual flexibility and strength. Improper form or weak links in the kinetic chain can result in poor performance or shoulder injuries, one of the most common injuries found in professional tennis.&nbsp;“Serving begins with the power of the feet driving into the ground, the kinetic energy then travels through the abdominal muscles, into the shoulder, the elbow, and finally, the hand. Somebody that doesn’t have an efficient or full kinetic chain may attempt the majority of the movement with the shoulder,” he said. “Let’s say there’s no flexibility in an athlete’s hips and ankles for the overhead squat. Power gets lost at these joints and doesn’t get transferred. The subject is then required to produce all of the strength with their pecs and shoulders. You might be surprised to see some of these elite athletes; some of them can’t hold a bar and descend without collapsing at the hips or lifting heels off the ground,” he continued.&nbsp;<img width="500" alt="_1080893" src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjEyNTMzMzg0MDY2LTE2MTI1MzMzODQwNjYucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" srcset="https://www.xsens.com/hs-fs/hubfs/_1080893.jpg?width=250&amp;name=_1080893.jpg 250w, https://www.xsens.com/hs-fs/hubfs/_1080893.jpg?width=500&amp;name=_1080893.jpg 500w, https://www.xsens.com/hs-fs/hubfs/_1080893.jpg?width=750&amp;name=_1080893.jpg 750w, https://www.xsens.com/hs-fs/hubfs/_1080893.jpg?width=1000&amp;name=_1080893.jpg 1000w, https://www.xsens.com/hs-fs/hubfs/_1080893.jpg?width=1250&amp;name=_1080893.jpg 1250w, https://www.xsens.com/hs-fs/hubfs/_1080893.jpg?width=1500&amp;name=_1080893.jpg 1500w" sizes="(max-width: 500px) 100vw, 500px" class="fr-fic fr-dii" style="width: 778px;">The precision and durability of the robust Xsens sensors make them the perfect solution for professional tennis analysis, and during his research, the sensors provided a means to accurately assess and compare techniques.&nbsp;“If you then compare a well-performed serve to a less accurate or powerful serve with different movement patterns, you can show they’re exposing themselves to more injuries. This could be mitigated through strengthening of the rotator cuff, opening hips, or strengthening the posterior chain to explode upwards and come down on the ball. By looking at the basic movements using Xsens data, you can piece together the bigger movements one joint at a time,” explained Dr. Moorhead.&nbsp;Dr. Moorhead was able to utilize the Xsens motion tracking alongside a high-speed video that matched Xsens’ 240 samples per second. This meant he could show how the athletes moved on camera and then relay it with the exact data.&nbsp;“Athletes are not scientists, so if you can present a graph next to a video then that can make a huge difference. Here, you can see an example of an athlete which performed an Overhead squat. By looking at the maximal range of motion on each repetition, it is apparent that the athlete has diminished flexibility in his right lower extremity. This could be due to injury, or lack of training.”<img width="1000" src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjEyNTMzMzg0NDczLTE2MTI1MzMzODQ0NzMucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" class="fr-fic fr-dii">“The Xsens avatar also comes in great use for providing a visual aid when we’re discussing the findings with the athletes,” said Dr. Moorhead. “While athletes can watch a video, seeing an up-close, virtual model of their exact movement which can be rotated, stopped, and replayed, really helps the athlete to perceive what happens during each movement and understand where they can improve,” he added.<img width="1000" src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjEyNTMzMzg1NTMwLTE2MTI1MzMzODU1MzAucG5nIiwiZWRpdHMiOnsicmVzaXplIjp7IndpZHRoIjo5NTAsImZpdCI6ImNvdmVyIn19fQ==" class="fr-fic fr-dii">Benchmarking a tennis seasonSo far, Dr. Moorhead has analyzed the first part of the junior athlete’s tennis season, tracking large quantities of biomechanical data. While this data provides an in-depth look into the performance of the athletes, with significant adaptations already being made to technique and performance, the long-term analysis of an entire tennis season will help benchmark performance and create databases for reference.&nbsp;“This is only the beginning, we want to build a database of normative values. For instance, each sport can have an age range and performance level with data that allows us to predict whether someone may be heading towards an injury or dip in form. We want to avoid injuries, but if they occur, we can use this data to help them recover faster too. We can use this data to improve avoidance of injury in other athletes due to the depth of data and length of time the data was recorded,” explained Dr. Moorhead. &nbsp;“This is based in hard science and we can significantly improve performance with this data.”<img width="5184" alt="_1080772" src="https://images.wevolver.com/eyJidWNrZXQiOiJ3ZXZvbHZlci1wcm9qZWN0LWltYWdlcyIsImtleSI6ImZyb2FsYS8xNjEyNTMzMzg3MDk0LTE2MTI1MzMzODcwOTQuanBlZyIsImVkaXRzIjp7InJlc2l6ZSI6eyJ3aWR0aCI6OTUwLCJmaXQiOiJjb3ZlciJ9fX0=" srcset="https://www.xsens.com/hs-fs/hubfs/_1080772.jpg?width=2592&amp;name=_1080772.jpg 2592w, https://www.xsens.com/hs-fs/hubfs/_1080772.jpg?width=5184&amp;name=_1080772.jpg 5184w, https://www.xsens.com/hs-fs/hubfs/_1080772.jpg?width=7776&amp;name=_1080772.jpg 7776w, https://www.xsens.com/hs-fs/hubfs/_1080772.jpg?width=10368&amp;name=_1080772.jpg 10368w, https://www.xsens.com/hs-fs/hubfs/_1080772.jpg?width=12960&amp;name=_1080772.jpg 12960w, https://www.xsens.com/hs-fs/hubfs/_1080772.jpg?width=15552&amp;name=_1080772.jpg 15552w" sizes="(max-width: 5184px) 100vw, 5184px" class="fr-fic fr-dii">Learn more about Xsens MVN AnalyzeFind out more about how you can utilize MVN Analyze to enhance your sports and biomechanical research <a href="https://www.xsens.com/products/mvn-analyze" target="_blank">here</a>.&nbsp;

with Xsens MVN Analyze

Enhancing the performance of Junior Pro Tennis Athletes

The hunt for viable solutions to help stop the spread of the COVID-19 pandemic now occupies the minds of almost every single profession. From engineers to mathematicians, and even to clothing brands who now make masks, everyone wants to do their part in stopping the spread. Po-Shen Loh, a Carnegie Mellon University math professor, is no different. He realized back in March that, with a combination of math and physics, he could develop a digital solution to stop the spread of COVID-19. The key initial insight was that COVID-19 spread uncontrollably because people infected others before they even felt the slightest bit sick, and his research area of network theory could be a game changer in pandemic response.NOVID is the product of two months of sleepless nights and passionate efforts of a group of around 20 individuals, many of whom are alumni and students of Carnegie Mellon University. They ultimately created the first anonymous contact tracing app published in the USA for both Android and iOS which did not collect any personal information (no GPS, phone number, email, name, etc).<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1589478851037-tech-1.png" style="width: 1080px;" class="fr-fic fr-dib"> Under the hood, what sets NOVID apart from other contact tracing applications is its use of ultrasonic technology to augment the wireless communication properties of Bluetooth. Many other proposed apps rely only on Bluetooth, but those signals pass through walls and ceilings in apartments and offices. Consequently, Bluetooth-only solutions have the dangerous potential of falsely detecting invalid contacts who had no way of transmitting the disease.Indeed, for ordinary applications the range of Bluetooth is an asset, but this technology does not produce precise short-range distance measurements. Other apps measure and record the received signal strength indicator (RSSI) corresponding to nearby devices, but they do not immediately calculate distance because it is still an unsolved problem to reliably translate RSSI into precise distance, especially in complex environments with obstacles that absorb the signal. Contact tracing is a very sensitive task which needs to be carefully executed, and thus the technology that people put in their pockets needs to be accurate. By leveraging ultrasonic technology, NOVID has developed a simple protocol which employs the physics of sound to accurately gauge distances between devices, with minimal impact on battery, and maximum privacy.&nbsp;<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1589478868207-tech-2.png" style="width: 1080px;" class="fr-fic fr-dib">NOVID uses the physics of sound to achieve far better precision than Bluetooth or GPS solutions, drastically reducing inaccurate detection. Sound travels at about 343 meters per second. If Device A sends a sound out at Time X, and Device B hears that sound at Time Y, the distance between them is a very simple calculation: Distance = Rate &times; Time, or 343 &times; (Time Y &minus; Time X). Of course, if people are close by, we are talking about milliseconds of difference between Time Y and Time X. NOVID has been successfully able to measure these differences.While using NOVID, the user&#39;s mobile device will emit an inaudible sound, a fraction of a second long, in the frequency range of 18&ndash;20 KHz. When two devices communicate, they are able to identify the distance of said interaction, and store it in their own local databases. Since the application is specifically programmed to work on these high frequencies, it will never process sounds even remotely close to the human voice frequency band. NOVID does not use GPS tracking, which means that all interactions are simply recorded on a relative distance basis, rather than on an absolute location basis. Instead of sensing that Device X and Device Y were both at Location Z, NOVID senses that the devices were simply Distance W from each other. To alert people who might have been infected by others, all that is needed is relative distance, not absolute location. The beauty of collecting only distance measurements is that they avoid recording or storing any location data, which is inherently non-private. (Your typical location at 3 a.m. indicates a lot about you.)To ensure privacy, each user is only assigned a randomly generated ID at install time, rather than a username. Without any personal details, the app then sends random temporary ID&rsquo;s to nearby devices, which they record as interactions. When an ill user self-reports an anonymous positive COVID-19 test, we process all of that person&#39;s prior interactions, and then send out anonymized alerts to all of the possibly exposed users. NOVID was founded by a network theory expert, and as such, employs efficient network algorithms to be able to not only alert people to first-order (direct contact) exposures, but also second-order, and even third-order connections to the virus.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1589478885909-tech-3.png" style="width: 1080px;" class="fr-fic fr-dib"> The user experience of the application has always been at the forefront of NOVID. As much as contact tracing is a technical challenge, it is first-and-foremost a social challenge. Without a good user experience, the technical capabilities will not fulfill their potential, and vice-versa. NOVID allows users to view informative interaction metrics displayed on the dashboard; information empowers users to take control of their actions, in a time where we have very little control over the world around us.<img src="https://s3-us-west-1.amazonaws.com/wevolver-project-images/froala%2F1589478971313-novid-mockups1.png" style="width: 1080px;" class="fr-fic fr-dib"><iframe width="640" height="360" src="https://www.youtube.com/embed/o4gZxI0Zkqs?&feature=emb_title&wmode=opaque" frameborder="0" allowfullscreen="" class="fr-draggable"></iframe> The dashboard has three main purposes: self-reporting, daily interaction metrics, and possible exposure alerts. The Self-Report section allows users to report whether or not they have tested positive for COVID-19. The Daily Interactions section features a chart that displays how many unique interactions a user has had each day over the past week, along with how many of those interactions lasted more than 15 minutes. Finally, the Possible Exposures section displays possible transmission events between users, broken down by the aforementioned degrees of contact: direct contact, indirect contact, and distant contact.&nbsp;Empowering users to know when they&rsquo;ve had exposures allows them to make informed choices about social distancing, thus protecting the rest of their community from further exposures. NOVID is beyond just a contact tracing application&mdash;it is a movement towards a safer global reopening.NOVID is currently available for download on both the iOS and Android App stores. Please download to help <a href="https://www.novid.org" target="_blank">Stop the Spread</a>.

sensors

ORGANIZATIONS. SHAPING THE INDUSTRY.

Xsens

Disruptive Technologies

Imagimob

Synaptics

AIRSKIN

Latest Posts

Sensor Fusion: Everything You Need to Know

Unique sensor network for measuring greenhouse gases

Voltage from wood

Podcast: Diagnosing Alzheimer's At A Glance, The Stress You Sweat, Bloom's Better Bioplastics

Enhancing the performance of Junior Pro Tennis Athletes

NOVID: Anonymous Contact Tracing with Ultrasound

Industry 4.0 Deep Dive. Part 3: Networks and Sensors