Erik: Welcome to the Industrial IoT Spotlight, your number one spot for insight from industrial IoT thought leaders who are transforming businesses today with your host, Erik Walenza.
Welcome back to the Industrial IoT Spotlight podcast. I'm joined today by Vinay Nathan. Vinay is CEO and cofounder of Altizon, which is based out of the Bay Area, in Pune, India. Altizon is a company that focuses on the industrial Internet, and specifically on sensor data applications in SDKs. In your experience, have third party organizations, so for example, the OEM equipment manufacturers or supply chain partners or customers, have they been highly engaged stakeholders?
And I asked because in some cases, certainly, the OEMs, for example, have a stake in the data that's coming off their machines, they would like this, certainly for their R&D or their manufacturing operations supply chain partners or customers would certainly appreciate visibility in some cases into operations so that they can improve the performance of their organizations or visibility into, for example, when product will be arriving from a customer standpoint. So is this typically something that would be a consideration in a project? Or do you focus primarily on solving the primary objective which is often related to productivity and then these would be potential, add on case uses of the data but maybe something that you're not directly engaged in?
Vinay: So in the smart manufacturing use cases, the primary objectives are laid down by the end users and enterprises and you kind of have to focus on fulfilling that. But to your point, there, obviously are a vested interest or stake in understanding what's happening. But it's important to draw that line in terms of saying if this is about smart manufacturing and digital transformation, the end user is the person who's going to call the shots.
In some cases, some of this data may be going back to the OEM, but that's a rarity. The OEMs, for us, at least from our perspective, the equipment manufacturers have been more relevant for the condition-based monitoring or remote asset kind of situations where a pump manufacturer wants to know how the pump is working on the field or sensor manufacturer that wants to know how their sensors are doing on the ground. So the asset performance management use cases are where we have found where there is a direct lockstep with the OEM and you're driving that. And very often that sale is also made to the OEM. So, at least that's what we have seen.
In the rest of the cases, the OEM is essentially an enabler. We, of course, interact with them to understand in some esoteric cases if there's a protocol that is not readily available, whether they can enable that provide support for that.
Erik: You mentioned earlier that you have this report tracking ROI by use case in industry, would you be able to give us a quick executive summary of some of the key findings that you found through your research in this?
Vinay: We captured four years of work that we are done around 62 are connected plants at that point of time that we have and typically, when you're looking at productivity quality use cases as being the primary drivers. We try to look at okay, what are the key impact metrics that are there, so let me take an industry like automotive to quickly walk through an [inaudible 03:48].
Just to your point earlier, this is available off the web. So take out industrial automotive. And out of all the plants that are connected, what we realized was 46% of the use cases that were enabled was around productivity improvement. And that was followed by condition based maintenance has been 18%. And energy was the other big one at 24%. So automobile, this was kind of the main use cases that the customers were looking at for an industrial IoT project.
And when you look at the ROI impact in terms of what is the payback period for a lot of this, so something like productivity improvement which is really core, we find that it is a great use case to start off with because in more than 65% of the cases, the ROI is within 6-12 months for a lot of this. So that is a great metric to work with. And it allows us to guide customers in saying, hey, man, if you are an automobile company, that's what our salesperson would talk to you about in terms of saying, hey, let's look at the productivity metrics that you have and let's look at what it takes to get up and running with that such that I'll give you an ROI in 12 months against what you're trying to do. So that's one example. I hope that illustrates the point. And we provide this similar data across five or six verticals against projects that we have done.
Erik: Do you see a widespread in terms of the success of companies that certainly could be related to their operations but also to their ability to implement or to actually get their shop floor to adopt the technology? Or do you find a fairly narrow spread out in terms of the actual ROI?
Vinay: The nature of the use cases change. If you follow the right use cases for the right verticals, you're in a better position. So let me give you an example of what I mean. So let's say discreet industry, like automotive which traditionally has not invested because of the high CapEx or this is thin margin business, there's a lot of human element involved in the process and all of that. So very often, a lot of the analytics layers or the historian kind of product lines, for example, have never found purchases in these markets.
So, in this kind of a segment, the automation and visibility provided at a price point such that ROI is available in 6-12 months is something that is hugely attractive to that right industry. So it's not probably environment, health and safety that clicks first if you go to discrete. Whereas something like that looks in that process, site chemicals and things like that where the downtime matters a lot more, the fact that you're able to correlate some of these data sets and be able to give early warnings or condition based on the condition of an equipment is a no brainer very often. So you have to look at getting the right use case in front of the right customer, so that tends to make life a lot easier.
The second side of it is now we're finding it's getting to the fifth year of our journey, we had to do a lot more evangelism and sit next to the executive and explain what the benefits are kind of stuff in the first couple of years, three years of our journey. Today, by and large, what we're finding is enterprises have at least a first level view of what they want And then you do have to do some more, but you're not starting from the basics. Pretty much across the board there’s a secular kind of trend to say the industrial IoT is important to me and I now need to understand what I need to start first. That tends to be more the question as opposed to should I be doing any of this?
Erik: What might make sense, I think we saw certainly in China but I think also globally a lot of situations earlier where there was significant top down pressure to adopt something and then probably the right use case was not being adopted because there was a lack of clarity in terms of what that right use case was. If we kind of put your particular technology to the side and think about what have you seen differentiate highly successful customers from less successful customers in terms of their internal decision making process, their execution process, what recommendation would you make to a customer as they start out on this journey in order to help increase the overall ROI of their implementation?
Vinay: I will go back to a point I mentioned earlier are the cross functional team. To pretty much every successful implementation that we have had has had an element of a cross functional team where we have paired a business unit person from the actual function with an IT person who's also equally committed to the whole process. Top down is a given, unless you have top down acceptance or push, it's always difficult to drive a new change thing down on the hierarchy. So you need that as a top. But even after doing that, getting the right combination of a cross functional team increases the probability of success multiple.
The second thing is this critical aspect of choosing the right project to start with, rather than going for Big Bang stuff. Very often it makes sense to start small to iterate, see if it is working for you at some level. You may not have the entire roadmap for the next three years chopped out when you start. You should have it for a limited timeframe. If you're unsure of which four projects to start off with, define smaller versions of those that you can get done in three months, and then make a pick of what you want to scale. [inaudible 10:00]. It’s something that is very common to the IT industry but not necessarily so in the context of industrial world.
So, that mindset of being able to experiment fast and thereby arrive at the right starting point for your journey goes a very long way after the cross functional team angle that I mentioned in terms of enhancing the probability of success.
Erik: What is the most exciting trend either in technology or in business that you've started to see emerge over the past months that you're excited to follow and strategically that you're asking your team to focus on in the coming 12 or 24 months?
Vinay: There’re two or three things that excite me in a big way. And this is one that is well and truly underway, the Super Intelligent Edge being able to drive distributed intelligence and this is vital for us to drive wider use case adoption in several verticals. So this is a here and now kind of activity. But still there's a lot of work for Apple to realize products that are working on the ground in this area. So, the Super Intelligent Edge is something that is given now that I believe in and we are putting a lot of resources and effort of making that work the right way.
The next level that is interesting and exciting to me is the Industrial Data Commons. So we, for example, are working with several corporates. And in India, actually, we're part of a government initiative along with corporates to create an industry data common for a certain set of clusters in automotive, in particular, where performance benchmarking data and data of machine utilization and things like that are available in obviously anonymized format, but available for people to understand data scientists to work on from multiple angles.
So creation of that data commons is something that is super interesting and exciting to me, because that is when you start deriving a higher order yield or gain from doing all of this digitization, making this data available for both OEMs to look at and the enterprises to look at best practices evolving from benchmarking and things like that. So the Data Common, we are seeing several such initiatives happen around the world. And Germany has a few already in place. And India, for example, we are involved in this one.
The third one is something that across the SaaS world is a transition that is happening in the sense that you're becoming a value as a service. You're not just selling subscriptions and IP licenses and things like that. At the end of the day, you're tying your outcome and your future to the fact that on a continuous basis in the enterprise you’re actually getting value from what you're doing.
So from a productivity perspective, I mentioned metrics like UI and things like that. You now are getting into situations till you're driving outcome based off for the end customer and saying, hey, man, I will drive your way to increase it by 15% in 12 months and maintain it at say 70% or 80% or 90% over the next three years. So that additional step which may not be in a consortium format more often than not will help wider adoption and also drive enterprises to more readily be accepting of these solutions.
And there's a beautiful book written called Value as a Service by the founder of Cooper. And I see a lot of that also happening in the sense that now we are working very closely with a system integrators on the business outcomes being deliberate about what we are in opposition to promise as an outcome and even taking some of that as a yield back to us.
So these are the three trends as I look at it that the next 12-24 months will take shape in a more aggressive manner. One is the Super Intelligent Edge, second has to do with the Industrial Data Commons and how much of that can have an impact on industries as a whole, and the third is to do with a Value as a Service objective around providing SaaS a productivity app to an end customer or a condition based monitoring app or in customer [inaudible 14:33].
Erik: So I was speaking with the CEO and founder of Turbine, which is building IoT exchange. Are you familiar with the company by chance?
Vinay: No, I haven't.
Erik: Turbine they're based out of the valley. They've been in operation are in stealth mode doing R&D for 2.5-3 years now and they just entered Alchemist accelerator, so I think we'll probably see more of them as they start to go into pilot phase now. But their ability in a data exchange right now focused on transportation data because that tends to be more publicly available. Then you have Iota, for example, which is more of a based on a blockchain, but also focused on enabling machine to machine exchanges.
One thing that comes up often as a challenge is that if you're looking, for example, at public transportation data, that data from an organizational standpoint is relatively easy to work with because it's already in the public domain. So it's just around tagging the metadata and structuring it and maybe building contracts to govern use around that. If you're dealing with manufacturing data, then the technology problems can be addressed, but the legal and contractual challenges are immense.
You've been mentioning a more government oriented approach. These guys are taking very much a free market approach. How do you see this evolving in terms of companies being willing to share their data for the greater good or by monetizing it? Because obviously, there's a ton of value that's being locked up right now, how do you see this emerging so that that value can be released into the market?
Vinay: If you look at a smart manufacturing report, for example, it is an example of how abstracted data or derived information can be shared across an entire vertical ROI equipment that get connected highlights of trends and things like that. That is something that we can offer as a data service, which we are providing right now in the form of a report. There are certain data service APIs that we already exposed to some of our system integrators for these higher order applications. So this is the free market approach where it's going to be driven around that kind of approach.
The second is where there are data commons being created that are public. For example, we support a project called watchyourpower.org. What this does is it gives a power quality and availability information at this point of time across about 600 villages across India, and then in other three other countries as well, Indonesia and a few others. If you can imagine what this unlocks, power quality, but the quality is where it gets interesting. Because now, for example, for manufacturing, if quality is not up to scratch, it has equipment failure issues, and things like that.
As you can imagine now, this is like a public funded project around just providing poor quality and availability. Somebody could take this data and build a more sophisticated application that is more focused on saying, hey, man, this kind of equipment and if you go and deploy in this part of the world, you're going to have these kinds of problems.
So this is a second kind of situation where it may be a public works version, government is not involved in this. It's actually privately funded. But the point is that the data is public. So the first was proprietary data, but available in an abstracted format, anonymized format, such that no legal implication is being created. Second is where it's a privately funded project for public data.
And the third is the government kind of approach, which I mentioned earlier where the idea here is to provide a tool bench or a workbench where pilots can get executed here. One is, of course, in increasing awareness. And second is a whole bunch of piloting infrastructure where the infrastructure is provided by the government or it's a public private initiative in combination. Everything that happens there is available for the community to consume, whoever is a subscriber there.
So these are the three different models that I see emerging across at least what we have had an experience working with.
Erik: I think this is a very, very fascinating aspect of the market that has huge potential, but really has not been developed as rapidly as platforms sensor technology and so forth, just because of the legal and organizational challenges on top of the technology challenges. We'll go ahead and put those in the show notes. Vinay, thank you so much for taking the time to talk with me today.
Vinay: I appreciate it a lot, Erik. And it has been a pleasure being participating in the conversation. Thank you.
Erik: Thanks for tuning in to another edition of the Industrial IoT Spotlight. Don't forget to follow us on Twitter at IoTONEHQ, and to check out our database of case studies on IoTone.com. If you have unique insight or a project deployment story to share, we'd love to feature you on a future edition. Write us at erik.walenza@IoTone.com.