Supply Chain
Technology Podcast

SUMMARY KEYWORDS

Operational efficiency, IoT technologies, automated inventory, legacy inventory tracking, RFID challenges, BLE technology, demand signal, supply chain, distributor network, consignment model, data analytics, technology integration, business model, cost reduction, pilot phase.

SPEAKERS

Rajesh Chauhan, Scott Hurley

Scott Hurley  00:08

Welcome to supply chain tech podcast with Roambee. In this episode, we speak with Rajesh Chauhan, an Enterprise Architect and Technology Strategist at Interstate Batteries. Rajesh shares with us his experience of driving an operational efficiency strategy using live sales velocity and automated inventory replenishment signals across his dealerships. He also talks about the challenges he faced with the legacy inventory tracking systems like RFID, the solutions he has considered or decided against, and finally, what worked out for him, and so thank you for joining Raj again, and today, we’ll be discussing the use of IoT technologies and digitalization technologies for the purposes of gaining operational efficiencies as the sort of goal. So I’m curious for you to explain to us what are some of the specific problems or challenges that you’re trying to solve for through the deployment of these types of technologies.

Rajesh Chauhan  01:07

You’re talking specifically in our use case, right? And I, before I jump to that answer, I think it will be good to know what we do and why exactly that operational efficiency is critical to us. So we are a supply chain a pure B to B supply chain company. We don’t manufacture anything. And also market presence is in an after market battery business, right? We have very limited source or of suppliers who can provide us our parents, predominantly there are three or four known companies, some of them you know already is like JCI, Johnson control, X Johnson control now companies, clarios, Exide. There are a couple of more small and big companies around there, but predominantly those two or three of them. So a limited supply chain suppliers. We have a network of 250 or distributors through which we operate and service around 250,000 dealer locations. So if you most of the OE car manufacturers, other than POS, BMW, GM and Ford, their service center, if you go there, those batteries are ours. It is private label for them. We control roughly around 20% of the market in the aftermarket. Now, when you look into our our business model, what is our existing business model? It’s a high touch business model. Reason is, these distributors who operate in our ecosystem, they are independent entities. Also the product is delivered more of in a, I would say, a lack of control way how the product so distributors, to large extent, buy the product from Interstate Batteries. The product gets serviced or supplied by the supplier, direct to the distributor. So there is a disconnect in this whole whole process. So we lose the site of where our product is, how much is there, as well as from an operational aspect, there are multiple handshakes and our business processes which exist. They’re very manual. Technologies, flaky, um, to any kind of real time inventory, data quality, so there are multiple, what’s called underlying issues, which also adds on to the to the operational efficiencies which we have in or which we cannot account for as of right now. I hope that answers your questions around around the operational efficiencies or requirement for that in our current

Scott Hurley  04:08

Thank you. So what are the impacts of some of those inefficiencies in terms of on your business that you’re trying to measure and improve?

Rajesh Chauhan  04:19

Perfect. Thank you for asking that question. Our biggest challenge is the cost to deliver the product. So from the time, if I’ll take one more step back, our business model, our business is a fixed cost business model, okay, also the product is seasonal in nature and does has a shelf life. So two things, the seasonality, fixed costs and and it is being a what’s called perishable product. So

Scott Hurley  04:57

I’m not, I’m not a car guy. I. Um, per se. So what is the seasonality aspect of batteries? When are they more likely to to move?

Rajesh Chauhan  05:06

So higher the temperature, or the lower the temperature, you will see more sale. And because all the batteries are built to kind of have a longer life, or a sustained life at a standard temperature. So seasonality plays a huge thing. And when you have your supply chain where there is a fixed supply, right, they don’t scale up and down, right, but our business scales up and down. So when you have that kind of scenario, we have to manage our inventory, making sure we have the right product right place at right time. And sometime time could be weeks, and it could be a need where you could be slow in reacting to those scenario, and you will lose heavy, large amount of your business to your competitors. Okay, so going back to our questions around about cost to operate our business, and I’ll, I’ll probably share an idea around we’ve been operating this business from last 70 years, and I believe the way we operate, so think about it. We procure the product from the supplier. Goes to distributors. Distributors have their own territory, which is geographical specified. The there are multiple drivers or so called. We call them RSM, route service managers. They load the truck with a fixed quantity of batteries, purely by their knowledge or the local knowledge. What the distributor has it, right? Nothing has been digitized. It is all somebody’s mind, and somebody knows about their dealers, their their routes and everything. They load 400 batteries, 500 batteries truck. They’ll make a stop at those 2530, locations, which they have to go look at the rack and find out what is being sold. Remember, they might not have visited this place in last one week, two weeks or maybe month, right? Just happened to be there, and that’s where the first thing they figured it out, what sale really happen. Our business model is a consignment business model. So the inventory which is there at the dealer location is really ours. So we only get paid when dealers sells it. The problem is we get to that visibility late in the time, by the time we see the visibility, there is every possibility, two things we just don’t know whether all the batteries, which are sold by the dealer were ours. And second, what dealers sold and when he sold, because we have no idea. No no integrations, where their point of sale solution and some of these John garages don’t even use any kind of technology that most of their work is done on paper, right? So a business strategy was or was looked into and thought. About a couple of years back, we brought in Bain, who did a great job for us and provide us what our inefficiencies looks like and why it does exactly and how much do we really cost to sell one battery, I came to around 13 to $16 per battery is what is our spend to service a customer. And that’s a huge things. And out of that, if I, if I eliminate other deficiencies around sales and and hard shot, there are new capabilities which you talk about purely from a operation, supply chain, Operation perspective. There was roughly around six to $7 which was, which was an expensive proposition, and that is where we targeting right now ourselves, how to how to reduce that footprint, cost, what type of solutions or technologies have you previously tried to deploy to gain these operational efficiencies.

Scott Hurley  09:19

And where or what type of limitations may have you encountered with those legacy technologies?

Rajesh Chauhan  09:27

So we have, I would say, not heavily invested into technologies to to eliminate these operational efficiencies. And I think to large extent, it has to do with our our technology ecosystem, our distributor operations, which is 250 different locations, runs through an application which is built on VB and access. Think about it, 90% of our transactions go through that, and it is not one copy. It is 200 50 copies of them, so they don’t talk to each other. So the technology is a significant barrier. When you have every distributor playing in their own ecosystem and they don’t collaborate and communicate effectively and efficiently. Doesn’t matter what kind of overlay we’re going to put it over it from a technology space does not help. Sure. I think four years back, I would say four to five years back, we invested heavily into RFID technology to find out is, is if we take our racks and make them a smarter act using RFID technology, whenever the product being put there and pull out, we know exactly what the demand looks like, and we did a small pilot at four or five different distribution distributor locations. Two things which didn’t materialize. One RFID as a radio signal, had tough time breaking lead. The battery consists of lead and acid, right? And you RFID just could not penetrate. So the complexity became technology challenge for time of it. And second is cost for us to get over the hurdle of lead, we had to use the active, active RFID as our solution. We just became too expensive to support our business case. So, so that was shut down until after the when, when the new business strategy and other made it, made a call saying that, hey, we’re going to go pilot something in this technology space and find out, how can we effectively drive demand signal, right? We started looking into multiple technologies, and that’s where we found how the BLE can help enable us, and obviously our relationship with Roambe we do have other vendors also as part of that that conversation, but definitely at this point in time, whatever we are doing with Roambee in our pilot and POC has given us at least good results, at least from a site perspective, that if we continue to mature, that it will give us what we need.

Scott Hurley  12:32

Great. So for the sake of people that may be listening to this afterwards, can you talk a little bit about how you are deploying these low energy Bluetooth technologies and how that’s actually helping you with operational efficiencies from what you’ve seen so far.

Rajesh Chauhan  12:48

Perfect. So our first use case, or I would say, a couple of use cases, around using a demand signal to reduce our operational cost as well as create operational efficiency. Was first, we only show up to a dealer location where the demand exists, so do not do the mail crowds, aka second, the cost to operate the trucks because you’re loading large amount of quantity of lead and you’re moving it through the streets, which is a hazardous material, right? So you needed a heavy duty trucks, and obviously the commercial drivers, so you can reduce the cost of operating just purely the truck delivery model. And third thing was, let me think third thing was ability to start looking into this operational aspect of it, from not a distributor to a dealer operation model, but maybe a real time moving inventory from from a tiered supply chain model perspective. And when I say tiered supply chain model means something what Amazon’s and and the Walmarts do it so effectively, the product comes to a fulfillment center or their distributor center. From a distribution center, it goes to the fulfillment center, it’s locally distributed. So we are looking to use the same technology to create new business opportunities, to create new business operation, if we have the right technology in this space, right so to answer your question, we are in a pilot phase right now in one of our distributorship, and we are already running two routes as a dynamic crowd. So demand signal for the two routes we. I say two routes, a route is associated to a geographical location. So within that geographical location, as soon as we see that there is a demand, that demand get translated back to the visits, and the visits get translated to what route we should go today, and every way we reach to those dealers where the demand is, and that has helped reduced the amount of batteries which we carry, as well as our dealers. What we found is they have the product which they need at any given point in time, as well as we don’t lose sale to our competitors, right?

Scott Hurley  15:45

So from a technical perspective, how is the system actually working in store to be able to monitor demand?

Rajesh Chauhan  15:52

Perfect. So the idea, the concept, was as old as the RFID concept, which he said, because of our business being a consignment model, we wanted a lease intrusion on the dealer’s operations. So we don’t want the Joe’s garages or the Toyota service center or a Firestone service station to be impacted from from their daily operation. So how to eliminate or to build this technology solution in their normal operational mode was the critical success, okay? So how to do that was in place of asking them to do five or six steps for helping us enable this technology, we got to stay saying that they have to simply take out the tag, which is sitting on the battery, and put it in a small bucket. Okay, so think about it. We tag all the batteries in our warehouse. That battery gets loaded on the truck, get put it on the shelf, and what do we call a retail location which is nothing but a shelf? So that’s a logical entity, saying that the shelf is nothing but it plays wherever products are right. And the dealer, as soon as he sells the battery as part of the process, which we did not ask them to do anything more or different than take out that tag and put it in a small bucket with the bucket is called Soul bucket, okay, in this ecosystem now the technology takes it over. Under the hood, it is all the technology which is saying that where the tag was, and because the asset is tagged to that BLE tag, the presence and absence of tags now starts telling me where my product is, and my rags have gone from a dumb rack to a smart tracks, right?

Scott Hurley  18:04

So battery sold, tag taking off because no longer present or is present in a different location. You now know that that has moved out of the store. And if you have certain threshold, then you can say, Okay, we’re going to bring in X number to replace.

Rajesh Chauhan  18:19

Perfect so the absence of the tags in a retail location says that something happened, that something happens get translated back using the data processing or the IoT analytics, saying that knowing what is the consignment level at that data location is knowing what has happened to these two tags, or these five tags which have been not considered sold, there is a demand. Remember, it’s not one to one replacement, because you just want to be effectively going to that dealer, just because we might have 10 of the same thing in a given leader location. So I’ll just give you an example. They might have five empty 34 and and I’m just giving you, I hope that is okay with you, the product names, five, empty 34 and five empty 30 fives, okay? Or I would say two empty 35 if they sell two of the empty 34 that does not translate that I got to go visit because he already has to be more, right. So the data analytics drive that outcome, which says that here is the demand signal, and this demand signal translates into a visit. And that optimization happens outside the IoT space, but as part of data analytics operations, and which gets served back to our transport management system, which says. That, hey, you need to make these many visits. This is the route today you’re going to take. This is the product. And here’s the dealers that you’re going to visit. Right?

Scott Hurley  20:09

And have you found in your testing so far that it to be a simple enough process that a majority of the dealers are or the stores themselves are actually following the process.

Rajesh Chauhan  20:24

So there has been a training which is which we go through as part of onboarding a given dealer to this ecosystem. We have around 200 at Yeah, 280 odd dealers in this space right now as part of this pilot. So as part of onboarding, they go through a training. They’ve been told what exactly they have to do. There has been some pushback in some cases, but more education around saying that, hey, we are not doing this thing for our operation, we want to make sure it is for your benefit. So that is the critical aspect of it, because you don’t want their consumers to come to a dealer location and they don’t have the right battery to sell them. So how do we do that thing? Is that little help of them moving the tag, which we keep giving them, analogy that you go to Best Buy or stores, right? There are magnetic tags expensive product when you’re selling it, right? It just becomes a routine. So initially it becomes a bottleneck in the senses. I have to do this one step, but through the muscle memory building and training and making sure that they recognize this step. It has become a fairly acceptable solution right now. I don’t see any any dealer complaining about it.

Scott Hurley  21:51

And conversely, I’m guessing they don’t want a lot of inventory in their store that takes up space that they don’t need at that point in time.

Rajesh Chauhan  21:58

That is, that is one thing. Thank you. I think that’s it. That’s one aspect which we found, and we are trying hard. So there was always a process in the place making sure we have the right set of skews and the depth at the rack. In fact, we have floated a new process, going live on June 15, which is going much deeper into taking account of COVID 19 and current scenarios which are happening. It is far more intelligent process right now to make sure when we get the demand signal, when it gets converted to where exactly visit has to happen and what products which we have to put it has expanded beyond just purely the demand signal. It also looking into the aspect scenarios, different escalation, what is the distribution of cost looks like that in zip code, where the consumer come from? So there are a lot of other data sets which gets added to saying this, demand signal.

Scott Hurley  23:03

Great. So you touched on it briefly at the beginning of our conversation, but in terms of your ambitions to expand the usage of tracking and monitoring IoT based technologies for visibility purposes. What are you thinking what sort of an ideal end to end system in terms of what you want to be able to see and how that would benefit you versus what you may be able to get today?

Rajesh Chauhan  23:36

Yeah, most of our our roadblocks today, outside the realm of IOT space, where we are working with Roambee and a partnership has been our operation systems, data quality. So there are some foundational gaping to our ecosystem. You cannot have a you cannot have this isolated operational environment, 250 of them. So you bring all our, what’s called all the players on the same ecosystem. So there is a technology shift, or so called our technology transformation is is put in the queue right now, and we hope once it is there and we have this integrated ERP with the capabilities around from order to cash, sales code to pricing, transport management, inventory management, warehouse management planning. So it’s a integrated platform where the demand signal which is originating from this IoT services in its offer. Met, can be consumed and processed in real time across all our 250 distributors, not only that, across all the fulfillment center business, new business operation model, which we’re talking about, which I have mentioned earlier, and now we have a clear picture of what product resides at what place, right, where exactly it need to be sourced from, where it need to be sent to, and a last mile delivery, like how you see in the in the land of Amazon’s right, I place an order and I get the product in between, everything which is happening is so organic and so automatic that there is least number of manual touch points which we expect in this future state. So IoT is one space of it, but you got to have a strong foundational operational platform as well, because I could get really good demand signals, but if I cannot apply it in our operational world, there’s no benefit out of it.

Scott Hurley  26:08

Absolutely, it’s a means to me, but on into itself.

Rajesh Chauhan  26:11

Exactly. And I think we are looking from that aspect that we want to see the maturity of the technology. Can this be leveled and taken to the next level and on every state stage of our existing pilot, what we are testing, we are evaluating explicitly around, hey, if we choose to do this thing, how it’s going to fit into this long term roadmap, or long term technology roadmap, which we are looking at, right?

Scott Hurley  26:43

Okay, any other topics you want to discuss?

Rajesh Chauhan  26:52

and I feel like I would like to say thank you to the Roambee team, and it’s a it’s a hard coincidence as well as here is something which I would like to if somebody is listening and hearing this information right, For them to be to do a successful IoT, look into your own space and see what is, what makes sense to you. There are 15 different ways you can enable IoT. You have to play by your strength. What you want to take over from a technology aspect of it and what and for what things you want to partner with somebody else. So a lesson learned during our recent stint with BLE, this is, I think I did not mention that we experimented with RFID. We’ve experimented with Bluetooth also, but off of our own Okay, without taking a help of a good partner in that process, we started designing our own BLE tags. We started designing our own BLE gateways and the whole IoT infrastructure, and we went that route deep almost for six to eight months, and we realized that that is not the strength of interest rate battery where we want to apply, and the business case, or the proposition initially, which was to support this technology. It from a financial aspect of it from an outcome versus the financial model. It got thrown out pretty soon, and we realized that is going to be very, very expensive for us to build this ecosystem of our own. We brought in bunch of partners, started engaging with them, asking them how they can help. And through that process, we realized that, hey, we are purely an IoT consumers. We are not the IoT, what’s called the technology ownership perspective. We should not be reinventing the wheel if it is already been being taken care by some partners, and they are far more mature in that space. Go partner with them. So we looked into Roambee purely from that perspective. We said that, hey, how can we offload our deficiencies or our capabilities gap and look into how wrong we can fit into this equation? Not pretty sure there are other partners in the market, but as a as a technologies, which is my job, as a technology strategy, as a as an architect, as an IoT architect, look into what fits your use case. A BLE may not be your best solution. There could be other solution. You may not you may need BLE with other technologies like the. Visual or pictures or RFID. It all depends on what fits your best for your use cases, and how do you want to translate and embed into your operational model. What I’ve found is mostly people will take the technology and force fit into your operation model. What that causes is a rift or a fraction in your operation model becomes very difficult to adopt and continue to grow, and it falls apart. So look into this. This what’s called a harmonious integration between your business operation model. And how do you bring the IoT, two of them together, because if they don’t talk and they causes friction to each other, both will fail. Absolutely. I hope that that helps.

Scott Hurley  30:50

No, that gives a really good perspective. Thank you for that. Because, you know, in the market today, even still, we see that there’s a good number of IoT implementations that wind up failing. I think that’s a that’s a pretty strong root cause as to why, because they’re not matching operational needs against the technology being deployed, right?

Rajesh Chauhan  31:10

And whatever IoT you put in place solution, you put in place, your business has to accept saying that they might have to make some operation changes as well. So your old way of doing business and new way of doing business might look significant different, in some cases, maybe slightly different, but make sure you account for it, not accounting for that, and not prepared for it is really business gets a push back and says that, hey, technology, you’re trying to push me to do things which we never did it. So make sure you have the right set of training. Set the right goals, right process in place your new operating model. Make sure you have a clear visibility from where the IoT starts to engage in this ecosystem, all the way where it gets out of the IoT realm, more into their operational system, and how the technology put together is one solve. It’s not multiple solve. So that’s another thing. People look into IoT as a solve, and their operational system has different solve. They all got to come together, right?

Scott Hurley  32:21

All right. Wonderful Raj, thank you very much for your time today.

Rajesh Chauhan  32:24

Hey, thank you. Thank you for wonderful time, and I hope you have good, good time in Chicago. You here next time in Dallas? Give me a call.

Scott Hurley  32:41

Yeah, absolutely. I’d love to be able to follow up and see how it’s progressing, and ask some of the users, or even, you know, the end users, of the technology on your end, and really make sure that it from a use perspective, that it’s doing what it needs to be doing, and how we can improve. Perfect.

Rajesh Chauhan  33:01

Thank you. Thank you very much. Take care. Take care. Bye.

Scott Mears  33:06

Hi. My name is Scott Mears, and I’m one of the hosts of the Supply Chain Tech Podcast. With Roambee on this podcast, we talked to supply chain heroes from around the world about everything, ranging from the disruptions related to supply chains, their personal experiences with tracking technologies, strategies to build resilience and much, much more. We already have some recommended videos for you to the side of me, and if any of this sounds interesting to you, do subscribe to our Youtube channel and hit the bell icon so you don’t miss another Roambee video. I’ll see you next time you.