We Talk IoT Podcast - Episode 38: Exploring the Power of Mesh Networks

We Talk IoT Podcast - Episode 38: Exploring the Power of Mesh Networks

Show Video

foreign [Music] mesh Network leads to a lower cost of communication because the same information can flow through multiple Pathways so the the resiliency that you get in terms of the information that travels across the network is much better through a multi-path network instead of a point-to-point communication welcome to Aflac silica's we talk iot we'll chat with innovators experts and business owners to learn how they are implementing iot and using data to create new business opportunities I am your host Stephanie Ruth Haider [Music] thank you today we are diving into the realm of wireless mesh networks the invisible threats that seamlessly connect the devices in our iot ecosystem they are resilient robust and expand automatically as more devices are added so they do play a pivotal role in shaping our connected world I am delighted to have caitan Joshi lead product manager at Panasonic industry Europe on the show and he will share with us some insights on how these networks work how what challenges um they have and maybe what a little bit about what what the future holds to this piece of the iot puzzle welcome cheat and it's really great to have you on the show how are you thanks a lot Roots I'm very pleased to be here I'm a regular listener so I couldn't be more happier to be with you today great before we start can you maybe introduce yourself and tell us what you do at Panasonic absolutely so I've spent nearly seven years at Panasonic industry Europe now at Panasonic we have the luxury of as a company doing many many things but at the end of the day the goal is to be a contributor to the betterment of society we do that in many different ways by the means of our products by the means of our Technologies trying to make the best products and Technologies for our customers in fact I'm actually very proud to be talking to you as a Panasonic member because we have now initiated a carbon net IMPACT program that's called Panasonic green impact it has a goal of reducing up to 300 million tons of carbon dioxide oxide emissions by 2050. and in near term by 2030 our goal would be to just be a net zero company in our own systems so to be a part of such a company is actually quite awesome I would say and one of the most important things in context of the topic that we are going to discuss today is that mesh Technologies form a critical part of the puzzle because of the way how they contribute in the next industry 4.0 in order to bring down carbon dioxide emissions for our listeners who are not as deep in the topic as you are can you explain a little bit the concept of mesh networks yeah absolutely so so uh when we say mesh networks let me uh start by saying that we are actually talking about wireless mesh networks today you can eventually think about a wired mesh Network as well but in 2023 the idea is to Go Wireless in order to save car asks and if you just take a step back and not just a single step back if you go back 130 years when Wireless Communications started out right so there was this the father of Wireless Communications as a guy called Jagdish Chandra Bose and when in 9 1894 he first demoed the first wireless communication he just basically rang a electric bill from across the hall yeah wow that made newspapers uh headlines all across the world right got to 130 years later when my daughter was barely 17 months old right so when she's not able to get see a wireless bill working she looks at me funny you know and that that's the progress that we have made with Wireless Communications in last uh these many years why do I say this because when you think about the kind of complexity that needs to be hidden inside Wireless coming communication between two points from a point-to-point communication that is already complex enough but now just imagine how this complexity compounds itself when there are tens hundreds thousands tens of thousands different nodes that need to be part of a singular Network and have to share data streams where information goes from one point to the other to the from in a format of many to many communication so uh this is something that is today being addressed by a different sort of mesh Technologies there's no one-size-fits-all for mesh communication you have different implementations that you've used different sort of file level technology such as Wi-Fi such as Bluetooth such as threads such as zigbee and many many proprietary implementations over such Technologies right why are wireless mesh networks so essential for the internet of things when we think about internet of things you would think about where things that are your everyday objects but the real scale at which the things need to be connected is actually in the industry in the Brownfield projects where there is existing infrastructure that has already been deployed and it needs to be updated it needs to be made smart and in such cases trying to lay out wiring is not something that you would find feasible neither in terms of the Manpower that it requires nor in terms of the total Capital cost expenditure this is where having a wireless wireless front end that automatically lays out and deploys an auto autonomous mesh that is self-healing self-repairing in nature leads to very high efficiencies in terms of getting the work done and what other benefits are there I just said it's um it's cheaper and it's more resilient meaning one device disconnects it will be re-routed to a different device right that's how it works are there any other benefits I mean the the benefits at the end of the day are the costs that that you save so when you have uh speaking from an information Theory point of view having a mesh Network leads to a lower cost of communication because same information can flow through multiple Pathways so the the resiliency that you get in terms of the information that travels across the network is much better through a multi-path network instead of a point-to-point communication so that is the biggest advantage to start with eventually as you go down deeper into the applications of the mesh Network the advantages come in form of let's say application specific benefits that you can derive out of it particular technology so for example a Bluetooth based mesh would be an excellent choice for something that you would deploy in commercial lighting kind of situations a Wi-Fi mesh on the other hand with a heightened complexity is excellent for extending the range of your already high Data throughput Network but which doesn't contain as many nodes so it's really delving down into the depth of the individual technology that you would understand what the benefits are oh okay so yeah let's let's go into that you already mentioned commercial lightning versus extending the reception of your Wireless in your smart home I suppose or in your home in general would you like to focus more on Bluetooth mesh first and we can discuss um what other use cases there are yeah yeah absolutely so uh the thing that's really good about Bluetooth is that it has penetrated the human consciousness much like Wi-Fi so it's it's something that you immediately identify of what happened was that uh when Bluetooth started out they were just doing point-to-point connection or one-to-many Connections in form of broadcast messages but eventually in 2017 they released a new specification that in and of itself is a standalone specification not part of the main Bluetooth as you know it it's called Bluetooth mesh this allows some implementation of a mesh that is very unique in the sense of how it creates and how messages propagate so typically when you think about messages going across a network you would think that they are being routed across through a routing table or something like this like in traditional networks what Bluetooth mesh does is that it manages a very intuitive managed flooding approach where much like a flood the message eventually propagates throughout the entire network and by carefully choosing the parameters within the Bluetooth mesh you are able to have a very very decent performance of where messages bounce across multiple nodes and reach to the end device and the applications thereof are very evident in high density mesh networks such as Commercial Lighting something that you would see in hypermarkets or on office floors where you have sometimes hundreds and many times thousands of nodes that need to be controlled in a sync synchronized manner the scale of Bluetooth mesh somehow allows it to reach even up to multiple thousands as I said the biggest Network that I am personally aware of is has somewhere under just under 5000 nodes in a single Network oh wow so that's quite awesome so that means when you say controllers I can in the case of lightning check if all the lights are working if they're on and if they're off if there is there any other um kind of information that is transported so lighting is a very curious industry I mean once you start thinking about lighting you think okay I just need to flick the light bulb on or off right but once you are introduced to the different kind of things that are going in the lighting industry in terms of ambient Lighting in terms of mood in terms of setting scenes there have been many many different companies that have sprung around in the sector who are leveraging the mesh Technologies I can probably say the example of kasambi it's a very very well-known brand uh in the lighting space who have actually created an ecosystem of lighting fixtures and Lighting drivers and whatnot and you just need to plug in play and be introduced to the world of awesome ambient lighting just to finish on the last point you could imagine the kind of savings that you would get out of automating a complete buildings lighting infrastructure so with a single smartphone or same simple application you're able to control a multi-level multi-floor setup similarly and that leads to very high gains in terms of the carbon savings similarly if you think about solar industry I mean the way the deployment needs to happen of this technology is something that's that's non-trivial in nature so the solar industry today faces many challenges in terms of land acquisition in terms of having the right kind of solar panels available Etc but at the end of it it's all about how to optimize the interaction of the solar panels with the sun itself okay and for that what happens is that imagine a solar farm there are many many different balance later in rows and columns and what have you these need to be tilted so that the during the course of the day they receive optimal sunlight right now one way to do is to lay out wiring for everything right that is certainly not ideal the other way to do it would be to lay out a wireless mesh Network to control that and a lot of vendors use zigbee as a protocol in order to deploy something at small to medium scale because this is where the uh the the size of networks today is on solar Farms you would have typically 100 to 200 solar trackers now if you were to scale that to a bigger operation this is where things start to become really interesting because the bigger solar packs now have 10 000 or 20 000 solar panels in wide swords of land which goes from let's say a kilometer Square kind of land area in such cases the amount of nodes that would be going into a single mesh Network increase the complexity of the deployment and it's certainly something that you would have to think twice about when using any mesh Network because when you start deploying for example let's take zigbee as an example here right so as zigbee needs multiple coordinators right and you would expect that the if I have enough coordinators inside thing they'll be taking care of their own nodes their own end devices and I'll have let's say more or less functional mesh but since this is routing table based and these are simple micro controllers at some point of time they find their limitations and the network performance goes down so this already shows you root that there is some complexity uh to a mesh when you start scaling it within an application segment so for that if you were to switch to a proprietary image that I can cite here for example wire pass these guys they have a proprietary solution that also runs on 2.4 gigs and sub gigs and now on 1.9 gig as well okay it

allows you to have a really really performing highly scalable mesh solution that can be deployed in dense very dense or let's say a very sparse networks depending on what sort of low level file you are using so you can think about covering something like a personal area network to a wide area or city-wide area network based on the mesh technology that you might be using we will take a short break stay with us we will be hearing from our August very shortly this podcast is brought to you by afnet silica the engineers of evolution we help you bring secure intelligent and connected products to Market if you want to learn more about us we have put information and links in this episode show notes and you can also connect with us on LinkedIn or fnet-silica.com that's avnet-s-i-l-i-c-a.com what other challenges are there you just mentioned the complexity once you try to scale and for from a smart home to a big solar Park what other challenges are there to make wireless mesh Network work properly so when you say work properly it has to be heard in the context of what the end application is okay so if today your end application requires a high latency communication because you're okay with a sensor that is reporting a passing wild animal in a bridge in a or something like that okay that's low latency communication or let's say it's just tracking animals through the means of mesh Network and you should be fine right but if you take the same technology and then deploy it to a network that requires millisecond level of latency so quite low latency networks then you would think twice about what you are going for right similarly other challenges might be about the actual throughput then Network throughput that you want to achieve out of the whole mesh when information propagates from one end across multiple nodes What's the total through code that you would require and in that you would think about uh the different kind of topologies that you might adopt based on the technology that you're using the kind of support infrastructure that you might want to deploy alongside your mesh solution for example a Gateway routers or something like that so A reconfigurable or configurable solution depending on which technology you move ahead with and all of these challenges put together if you have a decent set of Technologies at your disposal you can eventually come up with a solution or solutions that work for one case or the other when you mentioned applications that require a very low latency what would be a good example for this a typical example of this could be I don't know EV charging okay right so uh so EV charging stations I mean it's it's a trend that has just picked up like crazy so uh first of all it started with residential charging and let's say industrial charges two per parking lot three per parking lot and then now what you see is that every uh supermarket has a dedicated parking lot with only charges deployed everywhere right so while there are let's say some time sensitive Notions that typically would not be well at risked by uh any sort of Wireless networking for that you would eventually go to a more reliable solution such as ethernet itself okay there are many use cases that may be uh well suited for wireless mesh networks the trend that we see here today in the market is that many many different vendors of EV charges are integrating sigpi in their Solutions as well as Wi-Fi okay this typically could be uh for exchanging usage data load balancing or all the different kind of informations that might be shared between two charges will the new standard meta in this case at some point yes so method is a very curious thing because matter is something that runs at an application Level and what it does it is is that it creates like a unified language for all of the smart objects to talk to each other right so as if today uh you can think of matter as being as it starts out in our time it's starting out making the everyday objects at home compatible with each other in the context of a smart home if you have an electric car it's essentially a smart computer with wheels on it right so uh I anticipate and all of us anticipate that very soon your EV charges and your smart cars will also be appearing on your home network and you could control them with your smart Google home or Alexa or whatever not so there are underlying protocols such as Wi-Fi or thread which is yet another form of mesh Network Mesh networking standard that from the basis for matter right okay so if your EV Charger needs to be identified on a smart home network having these underlying Technologies such as Wi-Fi or thread are actually quite important but I if I understand you correctly meta at this point is more of like a smart home use case and not in an industrial iot landscape so the guys at connectivity standard Alliance they have done a brilliant job so what they have focused on is right now how to make the fundamentals of having an ecosystem of products that talk to each other how to get that part right right so this would require a lot of details about digital identity on how to introduce a certain node in the network what are the different layers that could be used in the in the communication stack as I said Wi-Fi infrared what's a good end node with let's say sleepy Behavior which contains doesn't consume a lot of battery power what's on the other end something that's acting as like the central aggregator of information such as a Gateway or something they have done that but right and since the technology has been drafted the right way it's only a matter of time that this eventually finds more industrial use cases as well um in your opinion what other advances or improvements in wireless mesh technology might we see in the future I think mesh is not a new technology or uh to start with right so it has been around for a very long time what we see in the market right now is that there is adoption across various parts of the frequency spectrum historically Solutions such as zigbee mesh were already available in sub gigahertz Spectrum this is something that's very very interesting for very low throughput Communications for simple control applications such as switch or something like that but eventually once you have these mesh networks deployed at relatively high density and they are in a decent environment you would see a requirement of everything that is in context of a smart home or smart industry gradually climbing up to the 2.4 gigahertz ism find there are a lot

of solutions out there already some of them already cited because this allows you to package a lot of let's say not just the control information but a lot of information on the data payload itself so you can send a lot of sensory information across the network and the and that would lead to higher distributed intelligence use cases um and I think the last topic that comes to mind we haven't touched yet is uh security is that an issue that's a very good question Ruth and uh I think it's important to say that security in 2023 and on will be the Cornerstone of iot so uh based on everything that we are seeing today in terms of governments releasing regulations uh you would require in order to get your CE approval in Europe for example you need to have the Cyber resilience act that requires you to have certain level of security guarantees on your end iot product we've also also heard recently from FCC that they're talking about in context of these raising user awareness about security something like a five point energy rating system that they have on appliances they want to do something similar for the security as well and I think it is fair to say that all of these mesh technologies have baked some part of the security into them from the link level security to the the application package that you would for example deploy in case of a firmware update over mesh all of that can be encrypted in terms of the data package itself in terms of how you provision these devices if you take the example of matter application that you send over thread over Wi-Fi you need to have these pki infrastructures put in place you need to have the the root certificates that need to be issued by some certified productization Authority and all of these things so as a general Trend I would say that if we just only talk about the Technologies the security is baked into the link level already okay right so you would base on them and security as you would know is not like uh there's no single point of failure that can be tolerated in the system so Bluetooth or Wi-Fi links would bring their level of Securities on the thing then you would add on top of that encryption on the application Level and so on right and if our listeners are now trying to decide whether um Wireless Mesh networking is a good solution for them or their infrastructure and how I might want to get started with integrating them into their projects what would your advice be in general right the first thing from planning out an application is to really understand what's the end goal of your infrastructure right once you have identified the key differentiator that you want to deploy you would have to go along the lines of identifying the right technology that is suited for your own application and you only do that once you start thinking about the end application right secondly irrespective of which technology that you choose I'm a hardware guy right so I I would have to speak about the hardware itself you would have to really focus on a very very very good robust RF design because whatever you do how no matter how fancy your software mesh is all about bumping information through carefully designed nodes that all form part of a single Network and a network is nothing but let's say carefully tuned frequencies that and sharing that frequency in the best manner possible right so if your RF design doesn't isn't good enough if you're not able to design a radio that is centered at the intended frequency of communication then you would never derive the the fruits that these Technologies promise is there anything I have not asked you that you were missing the something that you would like to talk about I think we have had a very good discussion eventually I think uh we'll have to uh we'll have to reconvene in two years time Ruth I would say to see how the how this very fast moving space has changed which are the new application areas that have taken up we already see a lot of let's say smart metering applications asset monitoring asset tracking pipeline industry and you know forests that are getting these Technologies in terms of let's say preventing from a wildfire and stuff like that let's say let us give each other two years time frame and then reconvene then we can really talk about how the market has moved lovely I would say it's a date so talk to you again in two years hopefully in between that as well it was really fascinating and interesting thank you so much for your time Chetan and have a great rest of the day thanks a lot Ruth it was absolutely fantastic talking to you this was abnot sillycast we talk iot if you enjoyed this episode episode please subscribe and leave a rating talk to you soon [Music]

2023-08-31 22:05

Show Video

Other news