- 00:00 - Introduction
- 03:30 - Trends and Challenges in Transportation and Logistics Industry
- 07:22 - Power of Smart Transportation through industry 4.0, IIoT & AI
- 09:18 - Key Transportation Use Cases for IoT
- 15:32 - Addressing Data Connectivity and Availability in Transportation
- 22:11 - How MQTT-Based Broker can help
- 26:28 - Digital Reference Architecture
- 44:41 - Q&A
IoT technology is enabling Transportation organizations to build smarter fleet management systems that improve fleet optimization, driver safety, and fuel efficiency. Drones are also changing the way goods are transported, providing new modes of transportation. However, the challenge lies in providing real-time directional messaging and connectivity as vehicles move through unreliable cellular networks, leading to dropped connections and significant network latency.
This is where MQTT technology comes into play. Watch this webinar recording and learn how MQTT can:
Integrate bi-directional data flow with other enterprise systems for scheduling, dashboards, and supply chain systems.
Help maintain a persistent always-on connection between vehicles and the cloud
Guarantee reliable data delivery, enable secure non-addressable clients to prevent cyber-attacks
Are you a solution or IT architect looking to revolutionize your organization’s transportation and logistics operations? Look no further! This webinar will explore how IoT and MQTT technology are game changers in the transportation and logistics industry. We show reference architectures to help you take your logistics operations to the next level.
The webinar focuses on unleashing the power of IoT and MQTT in the transportation and logistics industry. The speaker is Ravi Subramanyan, Director of Industry Solutions Manufacturing at HiveMQ.
Global challenges in the transportation and logistics industry include the need for speed and accuracy, fragile global supply chains, high costs, rising fuel prices, shortage of skilled workers, changing regulations, and the push for sustainability.
IoT can help improve regulatory compliance, operations, customer service, reduce mileage/transport costs, mitigate supply chain risks, and reduce emissions in transportation.
Key IoT use cases in transportation include: fleet management, public transit management, inventory management, asset utilization, and geofencing.
MQTT is a lightweight publish-subscribe messaging protocol well-suited for IoT connectivity in constrained environments with limited bandwidth.
HiveMQ offers an enterprise-grade MQTT broker to address scalability, security, reliability etc. HiveMQ is available for cloud and self-managed deployments and has integrations with many external systems.
Use cases enabled by HiveMQ's MQTT broker include real-time passenger information system for Munich public transit, real-time driver communication for a large transportation and logistics customer, and Automated Guided Vehicle connectivity and control to support a warehouse logistics use case.
Jayashree Hegde 00:00:08.463 Hello, everyone. Good morning, good afternoon, good evening. I'm Jayashree Hegde welcoming you all to our webinar on Unleashing the Power of IoT and MQTT in Transportation and Logistics. We have an exciting agenda ahead today. Allow me to introduce you all to our speaker for today, Ravi Subramanyan, Director of Industry Solutions Manufacturing at HiveMQ, who will be sharing valuable insights and practical examples that will demonstrate the transformative potential of MQTT in the transportation and logistics industry. Before we begin our presentation, I would like to share a couple of housekeeping pointers. Firstly, this webinar is being recorded, and we will share the recording in a follow-up email. Secondly, we encourage you all to submit your questions through the Q&A box, which is located at the bottom of your screen. We will address as many questions as possible during this Q&A session. Lastly, a downloadable copy of the slides will be provided to all the attendees after the webinar via a follow-up email. During Q&A, I will launch a poll. I request you all to participate. And throughout the webinar, I will be sharing important links to useful resources and link to our Contact form in the chat section. So do keep an eye on the chat box and feel free to interact with everyone. So now without further ado, I will hand it over to Ravi. Welcome, everyone.
Ravi Subramanyan: 00:01:36.360 Thank you, Jayashree. Thank you, everyone, for joining the session. Good morning, good afternoon, wherever you are in the world. I'm Ravi Subramanyan, as Jayashree mentioned. I'm the industry solutions manager for manufacturing, but I also kind of look into transportation and logistics because there's a lot of aspects to at least logistics within manufacturing. So I work with a lot of logistics companies as well. So here is my quick bio. But let's jump into the agenda here, right? So this is kind of what we'll talk about today. And we'll try to make it as interactive and as practical as possible because I think that's why you're here, right? You want to understand what is it about transportation and logistics industries where IoT can play a huge use case. What does data connectivity and availability look like within that space? What are some of the challenges around that? What is MQTT as a protocol or as a format, a data format? And how can a broker, a data broker based on MQTT help with that, right? We'll look at some reference architectures. We'll look at specific use cases that HiveMQ help customers within transportation and logistics. And then finally, we'll look at our customers and how they're actually using this product and what they're planning to do with it. And all across, yes, we will absolutely talk about transportation and logistics, but at the end of the day, a lot of these use cases are also valid for other industry verticals as well, like other aspects of manufacturing, pharma or chemical, or oil and gas or energy or others like auto manufacturing or other areas as well.
Ravi Subramanyan: 00:03:23.534 So we'll kind of talk about some of those, okay? Let's talk about smart transportation. And first, let's look at some of the global challenges that we see in transportation and logistics industries based on talking to our customers and based on our own research. Speed and accuracy continues to be an absolutely important thing. And as much as you can get to the speed along with accuracy, you are golden, right? That continues to be a very, very important aspect. Global supply chains are not as [inaudible] as it was imagined before. Now, this is across all industries, especially with the wars in the world that's happening and the COVID situation. I think the whole industry across the board has realized that what they thought was like a strong supply chain that they had is not as strong. And transportation and logistics industry is no exception to that. Costs continue to be a very, very important factor within transportation and logistics industries. And it is difficult to cut the cost given all the complexities that it has. Again, very similar to that, fuel prices are rising and that is impacting the economy. There's a shortage of skilled drivers and skilled workers within the space, like in many other industries. Again, government regulations, especially after COVID and other situations, have kind of been rampant. And that is really something that's challenging. And last but not least, there is a push towards sustainability. How can you go ahead and do your operations in a profitable way, but at the same time, how can you be sustainable, right? How can you be better to the environment, if you will, right?
Ravi Subramanyan: 00:05:15.267 So that is also a huge trend that we're seeing when it comes to digital technologies within smart transportation, right? So smart transportation is really the outcome of what you can do within transportation and logistics, right? If you want to obviously cut down cost or improve productivity or be more sustainable, that's the outcome, right? And these are the technologies that will enable you to get to that, right? So cloud technologies, obviously, like the AWSs, the Azures, and the Googles of the world, but other similar technologies as well, right? And one of the major applications that we see is digital threads, right? How can you kind of create a digital thread or a virtual thread of your entire kind of operations in the cloud or on enterprise that you can then use to view how the operations are happening and then tweak things without actually changing any process on the fly, right? AI, ML, artificial intelligence and machine learning — how can that help you kind of figure out how to better run your fleet, for example, or your transportation, your logistics, whatever be the use case. Similar to that — advanced analytics. How can you use the power of the data coming from all of the different systems and can crunch them into one location so that you can then use them to be able to do that? Smart sensors, obviously, is a mechanism to get the data from the different systems, factory systems, or other logistic systems, third-party systems to be able to feed into your advanced analytics. And last but not least, we talked about sustainability — how electric vehicles play a huge role, if you will, in this area.
Power of Smart Transportation Through Industry 4.0, IIoT & AI
Ravi Subramanyan: 00:07:00.426 And so why are people talking about this, right? I mean, the results are there to be seen. Using IoT and industry 4.0 and other technologies, there's a lot of value that fleet managers and other persona within this area are gaining, be it improved operations, reduction in accident rate, reduction in theft, cost savings, obviously, like less insurance claims, better use of the fleet, efficiency, reduction in idle time, better safety, better customer behavior or driver behavior, and greater manageability, right? I mean, I want to be able to make sure that I have a view on what I need to do today and tomorrow, right? So these are things that customers and personas are seeing. And that's why this is kind of something that is very, very important and being taken seriously. So just looking at the big impact, right? I mean, we looked at some of the problems and trends. So the impact of IoT is in all of these different areas, right — be it improved regulatory compliance by tracking all these hours of service and ensuring that personnel don't exceed that, improved operations. Now, obviously, some of these are kind of with the fleet, kind of like a connotation, but these can also apply to logistics and other areas within your space as well.
Key Transportation Use Cases for IoT
Ravi Subramanyan: 00:10:22.039 I also ensure that my company is complying with all the health and safety regulations of the regulatory bodies while being obviously optimal on the fuel consumption and the routes that I'm taking to make sure that, from an environmental perspective, I'm meeting my goals. So those are some of the big things from a fleet management perspective. Second is public transit management. Here, we would like to present a use case that we have with one of our customers. The Munich Transit Authority is using HiveMQ, for example, to be able to provide comprehensive, detailed information to passengers on the different aspects of their fleet, right, be it buses or trains or subways and other vehicles that they have to be able to provide real-time timetable on those things so customers can make decisions based on that. Hey, if there is a break in service because of some accident or something happened, that information is provided so that they can decide alternative routes, arrivals and departure times, again, real-time information, general news, and other information that they need to know of and other kinds of disruption delays and things like that. The key thing here is the real-time nature of the data that you can provide so customers can take detailed steps based on real-time information that is being provided to them.
Ravi Subramanyan: 00:11:51.704 The third is inventory management, smart supply chain. So for example, you have AGVs and other devices that are doing things from a pack-and-ship perspective, allowing these to be able to automatically track the inventory using IoT sensors that then ensures that you have better inventory distribution, you can service your customer better for e-commerce, ensure that you'll eliminate these manual scannings through technologies like barcode scannings or QR code scannings, things that are happening automatically as opposed to being happening manually, right? All of this needs information to flow through, and IoT provides that. Again, that combined with other pieces of information like your inventory data or weather data, traffic patterns, user patterns, and other information will help you be better in your end-to-end supply chain perspective, which also, from a manufacturing perspective, helps you to avoid any wastage that you might encounter through over-production or under-production, right, because you have that real-time information of inventory that's coming in and you know exactly where the demand is. So you know where exactly you need to produce, what to produce more, what to produce less based on all of the information that you have within your disposal. From an asset utilization perspective, again, we talked about the total cost of ownership. It's about ensuring that your shipments go out full so that you're more profitable, putting the right number of vehicles on the road to ensure that you improve your customer service and reduce the cost per ride, right, and feeding the data from your IoT sensors into backend systems so that you can track the real-time information on the vehicles. You can do capacity planning and load planning. And then the operators can ensure that they decide which vehicles need to be dispatched for what use cases based on conditions in real time, right?
Ravi Subramanyan: 00:13:48.171 Again, going back to the fleet management use case — this is about the optimal asset utilization aspect of it. Geofencing, again, like a use case for fleet management. Again, you're able to set these virtual boundaries around specific locations like your warehouses, your distribution centers, your factories, and other key destinations. And you're able to figure out as vehicles go in and out of these specific locations. You're able to alert so that you know, hey, a delivery is being made. A shipment has been picked up, for example, right? So then you can provide better ETAs, estimated times of arrivals, as well as delivery notifications, which these days has become a norm, right? And then, again, automatic notifications based on certain events ensure that the cargo doesn't leave specific areas. Thefts are rampant in fleets, so you can eliminate that by making sure that, hey, this vehicle shouldn't be leaving this premise, and so that could potentially be a theft or an equipment abuse situation that I can then take action on and correct it, right? And the last one is absolutely key where you could — a lot of times, you have high-value shipments that are arriving and nobody is there to pick it up in time, right? And so that good is just sitting there and it's wasted, right? So as opposed to that, if you know that a high-value shipment is arriving, you can ensure that you mobilize other assets and the staff like forklifts or like the personnel or refrigerated areas if it needs refrigeration, for example, to make sure that that is all available to process that shipment on time so that it's efficient.
Data Connectivity and Availability Challenges
Ravi Subramanyan: 00:15:31.230 Now, let's talk about some of the data connectivity and availability challenges that you see around transportation and logistics, right? So we would like to talk about this data maturity model where it's kind of like a triangle as you can see. Data collection is at the bedrock, at the bottom of it, the most important, the foundational aspect of it. But then you kind of take this data and convert it to information by organizing it into your big data lakes and you're digitizing the data. And then it gets synthesized into knowledge, and then that leads to artificial intelligence, and ultimately, into actionable wisdom or insights. That is what leads to digital maturity, right? Companies want to be at the top level where you achieve the digital maturity or you have knowledge and information. But the key thing that a lot of companies struggle with is the data collection and availability, right, because data is sitting in different silos. Data is in different formats. You have connectivity issues where your network may not be always optimized to be able to send data. And your bandwidth may not be always available depending on — if I'm a logistics company, I have seasonality, right, where suddenly my number of orders go up, so the bandwidth goes down, so then information doesn't flow, right? So those are some of the key challenges. And so from a data collection perspective, we talk about a value creation loop where data collection is not a one time and you're done. So it has to be — data has to be — and it's an infinite loop, right? There is a process where you gather data that is analyzed, and ultimately, that is then shared with some enterprise system that needs to take action on it. And ultimately that enterprise system needs to then talk into the backend, maybe to the field, to be able to share some information based on what it sees, right?
Ravi Subramanyan: 00:17:30.578 And that's that whole feedback loop that needs to be fed through a data collection mechanism. Now, let's look at some of the challenges in a little bit more detail. And we briefly talked about it, right? The first is unreliable cellular networks, right? Typically, these assets within transportation and logistics are connected through cellular like, for example, vehicles or other systems. Now, as these systems move through different coverage areas, they get hit with congestion and network blind spots. I was at an ITS European Congress event in Lisbon, Portugal last week. And one of the persons that talked about smart traffic lights, right? So they're working with the village or the city, and they supply smart traffic lights. And one of the big areas is these traffic lights are connected through 3G — or 5G rather. And sometimes they lose connectivity, right? And when they lose connectivity, they're able to collect the information and keep it, but they're not able to send it out, right? And that leads to information silo, right, or information breakage, right? So that is kind of a big issue where if connections are lost because of just drop situation or just high latency, right, which can cause lost messages and slow responses, that is a huge challenge. The second challenge is just using technologies that are not quite suitable for it. So you want to be able to send commands and data sets between your fleets of clients and the backend system.
Ravi Subramanyan: 00:19:13.172 And if you're using some web technologies such as HTTP that are unidirectional, which are kind of more for internet of humans, the World Wide Web, that's not quite suitable for the two-way communication that's needed and also broadcast messages from many clients based on real-time information. Sometimes you hit a snag when you use such technologies. The third one is just scalability, right? We talked about it briefly where it's not like the data needs or the bandwidth needs are always the same. So there is always demand, right? For example, when you talk about logistics, right, there are always seasons where the demand is high, like for trucks or for airlines or rush hour in public transit. And then there is low period, right? So unfortunately, a lot of the systems are not meant for the ups and downs of the scalability requirements, and maintaining connections during these times in a reliable manner is a big challenge. Looking at fleet security, right, and security across different systems, there is obviously a major concern, right? You don't want a situation where a bad actor can enter into a system unknown to you, right? So there has to be a trusted environment, and to ensure that there's always control rested within the authority that is maintaining the system, and to ensure that there are no rogue actors that come into the system to kind of start manipulating the data, right? And this is something that is across the board, not just in transportation and logistics, but across the board in manufacturing, in connected systems, mobility. It's a big, big issue and that needs to be addressed.
Ravi Subramanyan: 00:21:01.673 So monitoring and troubleshooting individual fleet clients. Now, if you have a lot of connections, right, it's tough to figure out what is that one connection where you are having issues, right? And how can I make sure that I'm able to find that one connection, diagnose it, and rectify it in a way that it can then function as expected because that one connection could then skew your overall data in a way that could bring your entire system down, or subsystem down. So you want to be able to monitor and troubleshoot connections that are giving you issues. And that sometimes because of connectivity, it could be a big challenge. Networking costs, right? I mean, this is another big thing. Obviously, if you're using cellular and other technologies, there is a substantial cost to be able to operate your system. And again, as the system kind of keeps going up and up, your costs are not going to come down, right? And especially as you kind of cut across multiple carrier networks, you have roaming and other charges that come into play. And so those things need to be factored in as well. Okay. So now we talked about the different challenges. Now let's get into MQTT as a technology, right? And then how can an MQTT data broker help address some of these connectivity challenges, okay?
How MQTT-Based Data Broker Can Help
Ravi Subramanyan: 00:22:26.531 So let's first talk about MQTT as a protocol or as a standard. MQTT started in the late '90s in oil and gas as a message protocol. In oil and gas, for Phillips 66, the key problem statement was that they have remote assets in remote oil wells, right, or in the ship, or somewhere kind of where connectivity is really poor. Bandwidth is nonexistent, and sending people out there to be able to do stuff is almost next to impossible. So you need to be able to monitor some of the operations remotely, and not to say other kinds of challenges that also go with that, right? So there needed to be — there was a need for a protocol where constant communication is not an option, right? I establish direct connectivity and then I keep polling for information is not an option because you don't have the bandwidth. The connection is choppy at best, right? Network connectivity is choppy at best. Bandwidth is limited. So how can I ensure that I'm still able to connect in that scenario? That's where MQTT was born as a publish/subscribe-based technology where if I'm a sender, right, I'm a client that is producing information, let's say I'm a remote oil well, when I have information to publish, I can publish based on when connectivity is available and when I have information to publish. And I can publish it in small chunks of 200 KB of data, right? And whenever connectivity is available, that data will go out and that data will be collected by a broker that is sitting in a common enterprise location. That broker then sends that information that I just published to any subscribers that need that information.
Ravi Subramanyan: 00:24:19.776 Maybe it is needed by a backend system to be able to monitor that piece of equipment, or it may need to go to a mobile device that has an application that needs that information to show things, maybe like an alerting based on that information, right? So those are the kinds of things. So it's a publish/subscribe-based technology where you have a bunch of publishers and a bunch of subscribers, and the roles can reverse. Your subscribers could be publishers, publishers could be subscribers as well. But it just creates a hub and spoke or a single source of truth-based model where everything goes to a single location. And then from that single location, it gets disseminated to other areas or other applications that need that information, right? So it's an event-driven architecture and it works very well in constrained environments as we talked about. And the important thing is that it's built on top of TCP/IP, which is kind of a networking backbone. And this is kind of a messaging layer on top of that. And it's very ideal for the industrial internet of things. And it's adopted rampantly in the transportation and logistics industry as well because of all the benefits it brings. And again, so it's an open standard. It is maintained by a foundation. And there are many members to that foundation. And HiveMQ, for example, is a member of the foundation. And our product is based on that. And they build capabilities on top of that, okay? And so why is MQTT that popular? We did a survey with a 500-plus sample size around different areas. One of the questions was: What do you consider as a strategic need to fulfill your IoT strategy — which protocol? And MQTT came up really high on top of HTTP and others, right?
Digital Reference Architecture
Ravi Subramanyan: 00:26:15.057 So that's kind of where you really see the popularity of MQTT going up, especially when it comes to connecting data from your remote location into your enterprise or to the cloud. So let's look at some digital reference architectures using MQTT now. So the transportation and logistics industries — the current way is to be, kind of, you have different systems that are talking directly to other applications that need information, right? So it kind of creates this spaghetti architecture where you have different connections that are crisscrossing. So it's not an efficient use of your bandwidth. And from a connectivity perspective, it could be a really costly proposition, right? From a TCO or a connectivity or a network cost perspective, it's not efficient at all. As opposed to that, what we have here is what we call a Unified Namespace, which is kind of a centralized single source of truth that is going to be maintaining all of the topic name structures on data that needs to be transmitted. Like for example, you have a train that is sending information, right? And there is a particular topic name structure for that information. And then you could have an analytics application that needs that information that can be subscribed to that same topic namespace. So you can ensure that, from a security perspective, there's no rogue actors there. The train sending the information is always only received by somebody that is subscribed and that is a legitimate receiver of that information, right? So that ensures that that information is not only secure, but it's also efficiently transmitted to somebody that needs that information, right? So that's kind of how the Unified Namespace works with a single source of truth and sharing information between publishers and subscribers.
Introduction to HiveMQ
Ravi Subramanyan: 00:28:04.745 And looking at it from an MQTT perspective, broker perspective, this is one way to implement a Unified Namespace where you could have the same clients that are publishing the information that's going through a broker, which is an instance of the Unified Namespace. And you could have other applications that are subscribed to the same broker that need that information and then that can share information with the other clients. The great thing here is it's always bidirectional, which means that you have publishers that could publish information, but they also could consume data. And all they need to do is be subscribed to certain topic namespaces or certain topics within the broker, and they will automatically get the updates by exception if there are any updates around that. So again, very efficient use of your bandwidth, efficient use from a connectivity perspective, and also overall information is shared very efficiently as well. Let's now talk about HiveMQ and what we do within HiveMQ. As mentioned, HiveMQ is an enterprise-grade MQTT broker. We were founded in 2012, right, outside of Munich in a place called Landshut. We are the movers of the data from connected devices. We move data back and forth from connected devices to the enterprise and back in a fast, efficient, and reliable manner. That's what we do. And we deal with the data. So anything that we need to do to the data to make sure that it's normalized, it's available, it has lineage, it is in a state that can be consumed, all of that we do, right? We don't do the end application with the data, but we just enable you to, allow you to, do stuff with the data, right?
Ravi Subramanyan: 00:29:55.990 So we have 160-plus employees and 180-plus customers that trust us with various different applications here, as a quick view. And I have a different slide that shows a much larger view of our customers. We are part of the OASIS Foundation that manages MQTT as an open source, and we help contribute to it and make sure — based on the feedback that we receive from our customers — that it's fed into the standard because we truly believe in kind of expanding the pie and allowing more players in the space. Sparkplug is another technology or a protocol or a framework on top of MQTT that is more for providing more data modeling, if you will. So if you have a subsystem that has information that is more complex and that subsystem information needs to be shared with other applications, MQTT kind of keeps it more open on how you share that information, whereas Sparkplug adds a little bit more rigor around the model, if you will, and then making sure that when a new subsystem is born or a subsystem is dead, that information can be shared with other systems that need to know. So Sparkplug is based on MQTT, and it complements it with all this other additional feature functionality. And so again, how does MQTT and HiveMQ help the transportation and logistics industries? Again, the key thing is maintaining a persistent and always-on connection between the vehicle and the cloud or vehicle and the enterprise, because not every use case needs to go to the cloud and we can support both.
Ravi Subramanyan: 00:31:34.333 The guarantee of the reliable data delivery between the vehicle and the cloud, which is absolutely important because if you are a driver that is looking for information to make a decision on the routes that you need to take and how you need to do things, you need to have a guaranteed way to have the data delivered to you and you need to be able to trust that, right? Again, ensuring that the data is highly secure and ensuring that clients or old clients don't come into the system and ensuring that you reduce the potential for cyber attacks, right? That is something key that we provide. The scalability of our solution is really something that is to be seen, right? We can scale up to 200 million concurrent connections, which obviously is huge, but that's kind of like the level that we can support. We can support anywhere between 0 and 200, right? The reliability of our communication, right? We talked about guaranteed data delivery. And we also have high availability to go with that, which means we ensure that if your system goes down for whatever reason, we can reauthenticate all of the clients that are connected to your system and we can make sure that customers don't see any blip in their service, right? So that's something that we can do through our clustering technology, which provides high availability, and again, the integration with all of the other systems. It's not just about your core systems, but it's also about bringing data from different other systems that needs to combine together to be able to provide more insights, more value around supply chain and other key use cases. That's what we do.
Ravi Subramanyan: 00:33:09.383 And so just as kind of like a view of our platform, we talked about a lot of these things already. So we can obviously — we bring in the data through all these different clients in the MQTT format into our broker. And the broker serves as that Unified Namespace. The broker can be hosted, either self-hosted by our customer on-premises or on the cloud. And on-premises, you have different technologies like these Kubernetes clusters, OpenShift; and on the cloud, AWS or Google or Azure. We also offer a managed service offering, which is called HiveMQ Cloud, where we manage a cloud instance on behalf of our customer, where they don't have to worry about how it's maintained and managed. We do it on their behalf. And that's something that other customers have started valuing as well. So we've touched upon some of these, the high availability, scalability, reliability. Observability is something that we also provide, where if anything goes wrong within your connection, you're not seeing data in your end application, you should have the tools to be able to debug it, right? And that's what observability provides to you. So our Control Center, for example, basically, is a user interface that provides to you all details about the clusters, what is coming in, what is going out, and which is on, which is off. And we are also able to take this data and output that through our extensions framework into external systems like, say, Datadog or other systems where you can have detailed dashboards that are showing how the system is performing.
Ravi Subramanyan: 00:34:39.323 We have another interesting extension called Distributed Tracing, which can allow you to put different traces all across the system to make sure that you're able to get to the root cause of what the problem is. That's around observability, right? Talking about extensions, this is kind of like basically an application extension to what we have within our MQTT platform. So we bring in the data in the MQTT format, but we recognize the fact that this data needs to be then taken to other applications and systems that may not be on the MQTT format, right? For example, streaming analytics, which is a very popular use case through Apache Kafka, Confluent being the end application. So we can translate the MQTT data into that format. If it is a pure cloud data that's needed, either on AWS or Azure or Google, so we have extensions that can send the data to those platforms and databases — that's a key aspect of what we do, translating the data and sending it to different databases. And the other thing that we do is what we call — in addition to all the security that we provide, we also have a Security Extension, which allows you to add additional security that IT can provide, right? Now, let's get into some cool architectures that we have helped out with our customers and partners. The German Association for Automotive Industry came to us and said, "Hey, we are kind of developing this architecture for AGVs and SCADA systems, right?" And the use case is obviously warehouse logistics. And we would like to use your system. So we work with them on this architecture so that they can be part of the specification, so that anybody that uses the specification can use our broker as a part of that, right?
Ravi Subramanyan: 00:36:30.783 And the key thing there was being able to clearly define your — be able to bring data from these different AGVs, if you will, into your enterprise, right, and combine the data with the other pieces of data from your control system, and then be able to centrally monitor it and authenticate it, provide any security on top of it, and enable bidirectional communication in a way that would allow you to share this information with the cloud or with the enterprise. So this is kind of the architecture for that, where we are showing the clustering technology with HiveMQ, where we are able to replicate the data that we have within a broker on different clusters. And then if one of the clusters goes down for whatever reason, the second cluster can automatically take over without any break in service. And from that point on, it will just service the main cluster while the other cluster that went down comes up, right? So it's very seamless that way. So from an end-customer perspective, they wouldn't see any breakage in their customer user interface, right? Things will be seamless from their perspective. The second is a very interesting use case that we have in architecture for a large transportation and logistics customer, where the need for accurate messaging to and from their drivers is absolutely important for them to complete their jobs, right? I mean, vehicles are fitted with mobile devices so that driver-to-driver communication and driver-to-back office communication can be enabled. And this whole backend is managed through a HiveMQ Cloud solution, and then that data then floats into either AWS or Azure based on the use case.
Ravi Subramanyan: 00:38:24.835 And then the key thing for them is uptime, right? I mean, uptime is the biggest need to be able to fulfill orders, and they would like to be able to enable text messaging and real-time data updates. Now, on the surface, it looks like a very simple use case: text messaging. It should be possible; why is there a big deal? This is kind of where, really, when the network is really pushed hard, right, when there is a slew of orders that comes in, right? In peak season during, say, Thanksgiving or Christmas or other major times where orders are coming in really high, this is where I think the messaging systems could get clogged. And with the previous solution, that's what happened. They were not able to figure out how to efficiently get the data out. They were losing orders because of that, or they weren't able to fulfill orders on time because of that. Once they moved to a HiveMQ solution based on MQTT, it just was seamless. So one of the feedbacks that they gave is that during the last season, everything else failed except for our solution. Our solution was the only one that could stand the brunt of all of the things that the systems were going through, right? So they were able to increase the uptime, and also they were able to accurately deliver goods through our architecture here, where we are able to provide them with the ability to steer traffic. We were able to provide the custom technologies, and we were able to scale up to millions of vehicles that are on the road that are transmitting data so that you can share information back and forth. So that's this use case. So this is a view of all of the extensions that we provide. We briefly talked about it. Let's talk about it again.
Ravi Subramanyan: 00:40:08.607 So your Kafka Extension has to be able to send and receive data to a Kafka messaging platform from MQTT, from HiveMQ. Enterprise Security Extension is an extension that allows IT teams within various organizations to use the same security that they provide for other IT systems like, for example, Active Directory or LDAP and things like that that they may have for other IT systems on the broker also. So the broker can be another piece of their IT ecosystem without them having to have different security measures depending on which systems that they're using. So it kind of enables that. The Bridge Extension is key, especially in manufacturing where you have the local factories, for example, or the local machines that need information to be collected through a local broker. And then that needs to be bridged to an enterprise broker. So this Bridge Extension allows to create a tunnel between one broker sitting on the edge to the broker sitting on the enterprise. So it can enable a seamless communication, reducing the network overhead and network complexities. The Pub/Sub is for Google. Distributed Tracing — we talked about for debug purposes. And then Kinesis is for Amazon. And we have a lot more that we're working on. And we provide an SDK that we allow customers to develop their own SDKs or own extensions as they see fit. Now, let's talk about some of our transportation logistics customers and then I'll leave some time for questions. We talked about the Munich Transit System. I'm not going to go back into it again, but this is kind of like their smart mobility.
Our Customers in Transportation and Logistics
Ravi Subramanyan: 00:41:55.402 They are deploying HiveMQ to be able to get real-time information on their passenger information system that they've created. The whole thing is fed through the real-time information coming in from HiveMQ. And the end result is that they are deployed with 500+ information monitors and 2000+ buses and tram stops across Munich. And then they are providing real-time information with constant monitoring. And the cost is low. And then the cost of implementation is low. And then maintenance is also low. Because MQTT has a very low overhead, it enables that. And this is another similar use case where this application from FELA, which is also a management systems company that specializes in smart scalable solutions for public transportation, where they wanted to be able to send the real-time timetable and tracking tool information into the vehicles to make sure that they have the most up-to-date information. And they are using HiveMQ Broker as a method for communication. With that, they're now able to have an event-driven way of getting data, which is more accurate and has less overhead. They don't have to poll for information. They get the information when data is available. And then the software versions and the timetables could be pushed to the vehicles as needed for the entire fleet. So it's kind of a — it could be pushed one time and every end vehicle could receive that and have the most up-to-date information, okay?
Ravi Subramanyan: 00:43:32.448 Let's look at our overall customers. So these are kind of our list of different customers. So obviously, apart from transportation and logistics, we are also into manufacturing. We are into connected cars. That's one of the areas that we started with, connected vehicles, connected devices, and anything that needs data to flow. Netflix, for example, right, is a customer on being able to test their backend. Any device that has the Netflix button needs to be tested before it's launched. That whole backend testing is done by us. A lot of connectivity providers, like carriers, use us. T-Mobile and others use us for ensuring that they're able to efficiently send and receive information from their enterprise into their various mobile devices. Before we go into Q&A, this is kind of like some of our resources. Obviously, Jayashree will share that with you as well. My contact information is here and you'll get this information if you want to reach back to me, if you have any questions or follow-ups that you need to do. But I'll leave it on this slide for you to ask any questions. And maybe, Jayashree, you can run the poll at this time.
Jayashree Hegde 00:44:49.120 Cool. Thank you so much, Ravi. This was a very insightful session. I am sure all the attendees have learned a lot from this session. So I will launch the poll real quick, and I will run it throughout the Q&A. I said that you all can cast your votes. I request you all to participate. So while the polls are running, maybe we can pick up the questions that are coming in, really interesting ones. So the first question is: What are the key things that are important for the transportation and logistics industry that MQTT is suitable for?
Ravi Subramanyan: 00:45:34.199 Yeah, yeah. And that's a great question. Thank you so much for asking that. And we touched upon it briefly as we went through the presentation, right? Obviously, at a high level, you have regulatory compliance issues, you have the big push to cut costs and be more efficient, be more sustainable, and also be able to have reliable messaging across, if you will, right, and then how you can do that in spite of your connectivity challenges that you face. We talked about some of that as we went through the different use cases and the challenges and how MQTT, being a publish/subscribe-based protocol, can help with that situation because it only transmits messages in small chunks. And it's event-based where if you have information, you publish it; and if you don't, you don't need to, right? So you don't waste a pipe, a connection pipe just because you're expecting some information down the line. It's exception-based. So information that is available and needs to be published gets published to a broker. And then the broker knows who needs to receive that information because that fact has already been established based on the topics and the name structures and things like that. And then that end application or system or the device gets that information.
Ravi Subramanyan: 00:46:57.001 For example, in the logistics case, if the drivers need to know some information for doing their routes, the fleet manager or whoever is in the backend sends that information, right? And once that information is available, any driver that is subscribed to that particular route gets that information so that they can take corrective actions on that. Or the other way is if the driver needs to communicate some information about, say, an accident or something to other drivers within their segment, they can do that, right? So not all drivers might need to know of that, right? Only a few drivers need to know based on the routes, they can do that. And that MQTT enables that to happen.
Jayashree Hegde 00:47:38.772 Thank you. Thank you so much, Ravi. There is a follow-up question from the same participant. He's asking, what does HiveMQ offer that sets it apart from others?
Ravi Subramanyan: 00:47:49.946 Yeah, yeah. And again, maybe I can also talk about what sets it apart, not only for transportation and logistics, but other industries as well, right? So MQTT is an open-source protocol that is managed by the Eclipse Foundation, and we are members of it. And it has clear use cases. We talked about it at length, right? HiveMQ has taken that MQTT protocol and created an enterprise-grade solution around that. Now, what does that enterprise-grade mean, right? Obviously, when you start off on your IoT journey, your number of connections is really small, the scalability requirements are small. Security requirements — they're generally small as well. So an open source broker, for example, Mosquitto or somebody else would be very, very ideal for that because you're just starting off and it's a DIY kind of situation. But when it comes to kind of scale up, right, when it comes to implementing it across your entire fleet, for example, or your entire logistics system or your entire manufacturing, say, plant floors across the board, say in pharma or in automotive manufacturing or chemical manufacturing, this is where you need that enterprise-grade solution. So HiveMQ, we also can start small. Obviously, we recognize that companies want to start small, but the thing is — we are able to provide the vision in terms of how you can easily go up to the enterprise or scale up, if you will, to the enterprise-grade that you need. Especially if you need high availability, you are in an industry where your customers shouldn't perceive any break in their use case, right?
Ravi Subramanyan: 00:49:32.833 For example, if whatever operation that some device is performing is key to ensuring that you're able to produce all of the goods that you need, for example, right, that information is needed by the driver to make sure that they're able to make their deliveries properly, high availability, reliability, accuracy is absolutely important. And that's what MQTT and HiveMQ specifically could provide because of all the additional feature functionality that we provide. And this is obviously great for transportation and logistics, but it's also great for other industries as well.
Jayashree Hegde 00:50:06.358 Yeah. Thanks, Ravi. So there is a question in the chat section that I think Mr. [inaudible] is asking: Can we get a trial version of the test to test new IoT solutions being developed by third parties?
Ravi Subramanyan: 00:50:22.772 You mean a trial version of the actual broker itself?
Jayashree Hegde 00:50:26.405 Yes, I guess so. That is what he's —
Ravi Subramanyan: 00:50:30.347 Yes. Yeah. So yes, absolutely, right? So one of the key use cases, a lot of customers or partners work with us to kind of be able to connect their IoT devices directly to our broker so that they can help manage that connectivity from their IoT devices to the cloud and they add a device management layer on top of that so they can have management of those devices and then they establish that communication back and forth. So that's a very common use case, and we can definitely enable that. We can help customers start off with a trial version. Our software is obviously downloadable. I think we provided a link to that in one of the slides. You can try it. There are some restrictions, but we can help you overcome that restrictions by — you can reach out to us if you want to — if you have a specific customer use case and you want to try it on, do what we call a POC, proof of concept with that customer, we can absolutely enable that by providing our software and providing consulting work with you.
Jayashree Hegde 00:51:30.667 We will share all these useful links in the follow-up email as well. There is another question in the chat section. Caesar is asking: Which IoT and other information is relevant to the energy and utility industry?
Ravi Subramanyan: 00:51:47.652 Yeah. So that's a great question. There's a lot of information that's reliable. One thing that's top of mind is Overall Equipment Efficiency. I mean, that's kind of the world that we are all living in. How efficient is my equipment, right? In the energy industry, if you're doing specific things to generate energy, some equipment is doing something, right? You want to make sure that that equipment is performing optimally, right, and that needs to be tracked. Historically, there's been mechanisms available to do that on a small scale. But in this world, where everything is kind of much bigger than what it was before, you need to be able to share that information with higher layers, to be able to do that automatically for you and be able to do some analytics based on that information, right? Things like how different aspects of my equipment are doing, and how to make sure that I can predict failure before it happens. When my equipment needs to be maintained, if you will, right? How to make sure that my operation that is running is running on as low cost as possible, but at the same time, how much CO2 am I saving, right? So again, being able to track all that information is absolutely key for the energy industry, of course, for oil and gas, for other industries as well. I mean, certain KPIs, key performance indicators are more important for certain industries, but all of these, if you look at it in a totality perspective, are absolutely key.
Ravi Subramanyan: 00:53:22.446 And the important thing is making the equipment smart by adding those smart sensors that are capturing information like pressure, temperature, displacement, and other pieces of information which individually may not be as useful, but when combined with other pieces of information — is when the magic happens. And then all that can be kind of taken into consideration into an analytics application, either on the edge, if immediate action needs to be taken care of, or it could go to the enterprise or the cloud where further processing could happen through various applications and use cases. And HiveMQ can obviously enable both those use cases.
Jayashree Hegde 00:54:01.230 Thanks, Ravi. Thank you so much. Let me move on to the next question. It's there in the chat. Aaron is asking: How do you ensure Unified Namespace not becoming the bottleneck or single point of failure?
Ravi Subramanyan: 00:54:17.690 Yeah. So that's a very, very interesting question, right? So a single source of truth could also be the single point of failure, right? And this is where — I mean, obviously systems fail, stuff happens, right? I mean, connectivity is lost. I mean, if you're especially depending on 5G connectivity or factory connectivity, sometimes things happen, right? Clusters go down, servers go down. And so that will impact your ability to populate data, right? The first thing would be to have a backup plan, right? If your connectivity goes down, how important is it to ensure that it comes back up again immediately, right? This is where that high availability really comes into play. And this is where ensuring that you have that reliability built-in through the high availability is important. Typically when we work with customers to put together solutions, we always go with multiple clusters, right? At least three at a minimum to make sure that if one cluster goes down for whatever reason, a second cluster can automatically take over. Now, behind the scenes, HiveMQ has a very efficient solution where the data is replicated across clusters and it automatically kind of hands over from one cluster to another cluster. And this new cluster that comes up can then reauthenticate all of the connections that are coming in. Imagine you have millions and millions of connections coming in that need to be re-authenticated, right? And that's kind of like the power that MQTT and HiveMQ provides, where it's able to quickly do that so that you can get all your clients up and running, connected to the enterprise and to the cloud, and so that you can start trusting the data, right? If you have breakage in the data, then you start kind of worrying about what happened, whether the data is more accurate. So to address that, we also have the ability to buffer, right?
Ravi Subramanyan: 00:56:06.190 So if the connectivity is lost, we can buffer the data. And that buffer size, you can set it to whatever you want, right? And then depending on that, we can buffer up to that level and then send the data in the sequence that it was received so that, again, when the connection is established and when you look at the historical data, for example, you don't see any breakage in the data. The timestamp is exactly what it was. And then from a real-time perspective, also, yes, there is probably a slight blip, but that blip is very minimal, right? You're able to kind of recover from that quickly and show the data as it should be seen. So some of the ways we ensure that your Unified Namespace is not a single source of failure and it's able to enable the connectivity across the different systems and offer observability tools and other methods to kinds of debug issues if it happens so that you can rebond and reestablish connection as soon as possible.
Jayashree Hegde 00:57:08.017 Thank you. Thank you so much, Ravi. There is one more question in the chat.
Ravi Subramanyan: 00:57:18.886 Yeah. Do you have examples — I see one question from Caesar. Do you have examples and would be available to work with us in the energy utility and mobile industry space? Absolutely, Caesar. We can certainly do that. We have and we can share some of the use cases as well. In the energy and utility space, we have a number of customers using us. Obviously, in the oil and gas and renewable space, we have a number of customers using us. Mobile industry, you just saw T-Mobile, right? I mean, the carrier T-Mobile is basically using us to be able to establish connections back and forth with a lot of the backend clients, if you will, that are the end devices. And you can imagine millions and millions of connections that are trying to authenticate and re-authenticate with the server that needs to be managed. And we do that in a very efficient way. And all of the things that I mentioned about connectivity loss and reestablishing connection, high availability absolutely holds good for the mobility industry specifically, right? Because if you're using the cell phone and suddenly you see things that are down, right, and then you don't — you shouldn't start perceiving that as an issue, right? It should try to rebond as soon as possible. That whole backend is done through us.
Jayashree Hegde 00:58:35.967 Good. Thank you so much, Ravi. There is one more question in the chat. Elvis is asking, kindly elaborate more on how shrinkage is avoided? I think he's referring to slide 15 on smart inventory management.
Ravi Subramanyan: 00:58:51.591 Right, right. Yeah. I think in general, right, I mean, inventory — so you have better information about the inventory coming in, right, and you have better information on the demand going out. So because you have different pieces of data, you can ensure that you have a good view of all of the inventory that you have. So you're not sitting in on excess inventory. And at the same time, you're not sitting in on low inventory to fulfill your orders. So it kind of becomes kind of a more real-time, plugged-in ecosystem of stuff coming in and stuff going out and everything around that, right? And that's kind of what we mean by preventing inventory shrinkage and allowing you to kind of fulfill demand on a more real-time basis, if you will.
Jayashree Hegde 00:59:39.893 Thank you. Thank you so much, Ravi. There is one more interesting question: I want to get digitally transformed. How and where do I get started?
Ravi Subramanyan: 00:59:49.146 Oh, boy. What a question. We only have one minute, right? So yeah, I think that's a great question, right? I think it's about setting the goals on what aspects of digital transformation are going to enable you or help you get to the next level. Once you determine the goals, identify which technologies can help you do that. Obviously, Industry 4.0 or whatever that technology is and what are the key aspects. How IoT can actually help you with that, that is the next step. And then from a data connectivity perspective, which broker would be the right broker to work with you? And yeah, so we do that in a consultative manner. We obviously don't do that in isolation. We actually have a team of partners that we work with and also the consulting companies that we work with as well, system integrators that can kind of help you put these systems together. We are working on reference architectures for different use cases that can kind of help you in this area as well in terms of what are the things that I can quickly put together that will help me get there and then how can I just replicate this across my organization. So that's a short answer for a pretty loaded question there.
Jayashree Hegde 01:01:03.837 Thank you so much, Ravi. This has been really useful. And we are at the top of the hour. So any closing lines from you, Ravi, before we —
Ravi Subramanyan: 01:01:15.385 Yeah. So the closing statement is that I think, obviously, digital transformation and data connectivity and data movement is a journey, right? It's not like you put this together once and you're done. So it needs a lot of interaction between senior management, and people need to be involved in the decision process. So I would urge each and every one of you to make sure that if you're the decision maker, please bring in the end people or the end folks that are going to be actually using the system and make them part of the decision process so that it's better adopted, right? They feel that it is created for them, right? And it's not like taking jobs away from them, but to complement what they're already doing is what we are always trying to do. That's the message that I would like to leave.
Jayashree Hegde 01:02:05.306 Awesome. Thank you so much, Ravi. And thanks to all the participants for tuning in. We hope you all enjoyed this session. Give us a thumbs up if you all liked it. And like we already shared in the beginning of the session, we will send a follow-up email with the recording and the presentation slides, also all the useful resources. And if you have any follow-up questions, do reach out to us. And our community forum is always there to help you out with any technical questions you have. And thanks again for tuning in. See you all next time. Take care. Bye-bye.
Ravi Subramanyan: 01:02:41.992 Bye-bye.
Ravi Subramanyan, Director of Industry Solutions, Manufacturing at HiveMQ, has extensive experience delivering high-quality products and services that have generated revenues and cost savings of over $10B for companies such as Motorola, GE, Bosch, and Weir. Ravi has successfully launched products, established branding, and created product advertisements and marketing campaigns for global and regional business teams.