AI in the Supply Chain: Cold Chain Technologies’ Ranjeet Banerjee
34:43 Share

Today, we're airing an episode produced by our friends at the Modern CTO Podcast, who were kind enough to have me on recently as a guest. We talked about the rise of generative AI, what it means to be successful with technology, and some considerations for leaders to think about as they shepherd technology implementation efforts. Find the Modern CTO Podcast on Apple Podcast, Spotify, or wherever you get your podcast.

AI is pretty hot these days, but how often do we think of AI keeping things cool? Find out today as we talk with Ranjit Banerjee, CEO of Coldchain Technologies. Welcome to Me, Myself and AI, a podcast on artificial intelligence in business. Each episode, we introduce you to someone innovating with AI. I'm Sam Ransbotham, professor of information systems at Boston College.

I'm also the guest editor for the AI and Business Strategy Big Idea program at MIT Sloan Management Review. And I'm Shervon Kodabande, senior partner with BCG, and I co-lead BCG's AI practice in North America.

And together, MIT, SMR, and BCG have been researching AI for five years, interviewing hundreds of practitioners and surveying thousands of companies on what it takes to build and to deploy and scale AI capabilities across the organization and really transform the way organizations operate. Today, we're talking with Ranjit Banerjee, currently the CEO of Coldchain Technologies. Ranjit, great to talk with you again. Welcome. Thanks, Sam. And thanks, Shravan. Hi, good to see you.

Our podcast is Me, Myself and AI. And we're particularly interested in individual stories about people working with artificial intelligence. And I said welcome again because I think we got to know Ranjit from his work at Becton Dickinson. And Shervin's known him long before that, I think. Ranjit helped us with our research a couple of years ago and we did a webinar together. But Ranjit, you've got a new role now. Can you tell us about your work with Cold Chain Technologies? So I joined Cold Chain Technologies in August of 2020.

And it's been a very exciting journey. What excited me about the company and what it does and the entire industry space it's in is the potential to really transform and address a key need as it comes to the life science industry. And that is the need for

Really, the simplest way I can say this is, how do you provide assurance when you are transporting drugs, biologics, vaccines over long distances, over complicated routes? How are you providing assurance that the product that is getting transported is meeting all the right conditions and ensuring safety and efficacy? That's the challenge statement that got me excited about cold chain technologies.

which is what this company does. So CCT or Cold Chain Technologies is the leader in thermal assurance packaging for life sciences. At the time when I was taking on the role, COVID was starting to become an important topic, a pandemic. And since then, we've also played a huge role in supporting the distribution of COVID vaccines.

That seems important to me, and I'm sure my own personal health has benefited from that. But what does artificial intelligence have to do with transporting cold stuff? If you think of where we want to go is go from providing visibility today, which is available to providing assurance that the drug, the vaccine or the biologic, when it moves from point A to point B, and point B being the last mile, being the recipient, the patient,

How do we assure that the product has reached in the right conditions? And that's where AI and other technologies come into play. So think of it this way. Today, we can monitor where the product is, where the drug is, what conditions it's seeing. But we want to know more. What we want to know is when will it reach the patient?

And when it reaches the patient, what's the probability that it'll be safe and effective? Or is there a chance that there'll be an excursion? So imagine now there is a drug being transported. It's stuck at an airport, right? A lot of drugs are air freighted, right? So it's stuck at the airport. Now you want to know, okay, there's a delay. Is that a meaningful delay?

And if so, what actions we need to take? When you start translating this from simple visibility to providing assurance, that's where AI comes in.

This probability is an interesting take on packaging and logistics. So if I understand correctly, one of the applications of how you're using AI is to figure out what's the likelihood that the item that's being transported is going to get to the recipient, not just in time and in place, but also intact or in usable form.

Is that right? Correct. And then let's say the probability is low, then what would you do? There would be an intervention. Exactly. Let me explain where we are right now. So we have launched our first path platform, which I'm calling Visibility. It tells you exactly where a vaccine is, what conditions it sees, or it could be also a drug or a biologic, a cell and gene therapy product.

where is the product and what conditions it's seeing right now? So that's what we have today. What we are working on, and that's going to be coming in pretty soon, is how do you now take the routes? We know exactly where it's going to be shipping. We know the ambient conditions. We could take in other factors like delays. There could be issues that could create a disruption. And then you take that data

combine that with the design of the packaging and say, is there enough thermal capacity that can take care of these disruptions or not? And if the answer is yes, there's enough thermal capacity, yes, the product is going to reach maybe six hours late, but it's going to be good. That's one implication. The other could be it's going to arrive one day late and still going to be good. That's got another implication. The third could be it's going to arrive one day late and it's not going to be good.

Imagine now, once you have that information, you can then take intervention steps as simple as alerting the recipient or setting another shipment, but you can do a lot better and more rather than waiting and losing time and more importantly, patient safety implications. This is quite fascinating because I have to imagine a lot of our audience is familiar with the

inherent complexities of just the typical logistics problem, which is something has to go from point A to point B and go through a global supply chain with lots of disruptions. And that's a multi-million parameter problem, which is a typical logistics problem. Now, Ranjit, you're talking about overlaying a whole bunch of other conditions about 10%,

temperature and delays and packaging integrity and a lot of other things that make this one or two or three orders of magnitude even more complex, I could imagine that's going to be a daunting problem for any human to solve, which is where AI comes in, I would assume. Exactly. And we are also realistic that this is going to be a multi-generational approach and you're

If you kind of step back and say technology always is an enabler to solve problems, it's not an end in itself. But the goal here is what I like calling how do you deliver assurance when you are transporting drugs and pharmaceuticals? How do you deliver assurance? Assurance as defined by will the drug be effective?

when it reaches its intended point of use or not and will it be on time? And it's really a whole new value chain creation.

You know, there was a phrase, and this obviously has not come from me. This phrase was used quite intensively in COVID vaccine program was people said, look, you can have vaccine, but vaccine does not translate into vaccinations, right? And a lot of folks much smarter than me figured that out and said that, but that is key.

You could do a tremendous amount of work in developing the vaccine or the drug or the biologic, manufacturing it. But how do you make sure that you are assuring it, what we call the mid-mile and the last mile? As we are creating the use cases, we try to always start with real problems and then back into technology. So as we are creating reuse cases, just imagine the drug arrives.

and you say, okay, there's been an excursion. Now, the next question is, is the excursion relevant or not? We could create algorithms there to say, okay, here's the manufacturer's data of what is allowable, what is not allowable. If there's an excursion, is the excursion relevant to the point where you have to discard the drug or you can automate a lot of these to improve efficiency and more importantly, improve patient safety.

You can also create compliance benefits because everything can then be converted into a record that creates the chain of custody of the product. So those are some of the things that we are working on.

To what degree does this process transcend your organization? Are you working with your upstream manufacturers to coordinate better? How localized are these improvements to your specific process? And how much can you involve your upstream and downstream partners in that process? So it's also a really good question, Sam. At this point, it's us more connecting with our own manufacturing plants into this network. But I do see...

our ability to pull the thread all the way to our suppliers. You can take this multiple layers because in the same context of providing assurance and along with assurance, it's a question of cost.

the best thing you can do is provide a high level of assurance at the lowest possible cost, right? And that's where beyond just our manufacturing plants, our suppliers and their suppliers, all that comes in, you can actually, using data technology and AI, you can build systems that are enabling that. Is the premise here, though, that prior to

use of these kinds of technologies in the cold chain logistics is the premise that there would be drugs that would be used that weren't effective and nobody even knew about it. Yes, this is actually a well-documented fact, but it's not probably well understood, I would say, but it's been documented through studies.

If you look at the normal waste associated with temperature control pharmaceutical shipments, it could be as high as 10 to 15%. For vaccines, which is a subset, vaccine is a subset of the entire family of pharmaceuticals, but vaccines, I've seen studies where you can have 20% waste on vaccine. 10, 15, 20% is not uncommon, even higher sometimes depending on where you're sending it. Now,

I think what happened was, while this was all known, I think what happened was with COVID, there was a spotlight on how important it is to get this right. And that was another reason I was excited about, here's an opportunity to really create some transformational change. The other thing that I'll say is, if you look at pharmaceutical research and innovation,

A lot of pharmaceutical research and innovation today is focused on large molecules as opposed to the small molecules in organics. They're both focused on large molecules. Typically, large molecules, cell and gene therapy products, these require careful condition monitoring.

Not just temperature. Some of these are even beyond temperature. But these are overall condition monitoring. And that has to be as much thought through as the discovery and the manufacture of these drugs and pharmaceuticals. I love the options this creates because I can even see scenarios that, you know, let's say you're not going to get to the original destination in time, but you can say, hey, it's pretty close to this other place that could use it right now.

I can't see really a human thinking through all that. So what was the process for doing this before you started to implement these artificial intelligence technologies? How did you figure these implications out beforehand? I personally am a big believer, and our team did this as well, is you kind of don't lead with technology. Lead with what are the challenges that are out there.

And then come back and say, what technologies can we use to address those challenges? So we started and we did this over the last year or so. We spent a lot of time talking to the stakeholders, understanding. We also had internal experts. So we started with our own kind of view on what the industry issues are. We went and talked to customers and we started creating these use cases. Each use case is a real life problem problem.

that technology can potentially solve. We didn't even ask what kind of technology yet. It's just like, okay, what's the problem first? Then what we did was we tried to rank them, bucket them together, make them more meaningful. And then from several use cases, we came to like three or four big use cases. And then we said, okay, how do we now solve for them

And that's where the role of technology comes in. It could be AI. We are looking at things like blockchain, where you need to get data across different stakeholders. So there'll be different types of technologies that address the problem. But that's kind of how we went about trying to understand what are the big challenges out there.

Shervin, that's coming up as a theme from lots of people we talked to, this idea of leading with the problem. Well, I was just going to say, there is a quote in our 2019 report from Ranjit on exactly that, back when you were at Beckton Dickinson. I'm paraphrasing, but it was something along the lines of, companies spend a lot of wasted time on

looking at technology where they don't really have a good appreciation of what the business value is and what the strategy is. And I do remember a very eloquent quote from Ranjit around, you got to get the strategy right of what are we trying to get done? You'll find the technologies, but first focus on what is it that you're trying to get done. Actually, you certainly got a good segue there by bringing up Becton Dickinson. How did you end up in this position? Tell us a little bit about your background. Like, for example, I know that

Actually, I'm not even looking at our producer, Allison, and sound engineer Dave right now, because I know they play bingo with how Shervin and I have to mention chemical engineering every single episode. And right now they're both checking their bingo cards for this. But we're all chemical engineers here. How did you come from that chemical engineering background through these steps to end up where you are today?

I really don't use much of my chemical engineering these days. I graduated from IIT Kanpur in India with chemical engineering. Then I joined Unilever. I used chemical engineering there, and I also got into leadership positions. And then, to be honest with you, I think the chemical engineering teaches you a certain discipline, certain process thinking that is really good. I do believe, though, I sincerely believe this, that

There's a certain disciplined process thinking that chemical engineers have. I've actually seen this in other chemical engineers as well, that kind of helps you think through big processes, break it down into each manageable chunk and

So there's probably some amount of background training that chemical engineering helps me. After that, I joined Beck & Dickinson BD in India. I moved with them from India to Singapore to the U.S. I was with BD for a long time. And then more recently, in August of last year, I moved to Cold Chain Technologies and took over as a CEO. Well, take us from chemical engineering then. How did you get interested in artificial intelligence and machine learning as technologies?

Personally, this is different people have different points of view. But my point of view is this is a whole new revolution that is happening where using technologies that we did not have before, we can solve so many challenges that we could not before. And these challenges are it's in health care, it's in financial, in fintech, it's in so many areas. It's in what we just discussed.

In every part of the world, in every business, in every organization, it's almost like you've got to step back and say, look, let's make a list of these big, hairy problems we have never been able to tackle for the last 20 years and kind of step back and say, what can we do? How can we tackle these now through the use of technology? It's really interesting because, Ranjit, you're also pointing out that you need two lenses, right? You need the first lens around technology.

What can technology do for the garden variety problems that any business has had? And then you also mentioned this notion of assurance is a radical new idea or a relatively new concept. So the second lens is, what are some problems or opportunities that we didn't even think about? And you need to really reimagine or really rethink some of your ideas

old beliefs around what's important or what could be done, because with the advent of the new technology, it allows you to actually conceive of new questions and new problems. And I can imagine, in fact, that's maybe a question to you, is this notion of assurance in supply chain, do you see that making a leap into more normal aspects of our transport that is sort of non-necessarily temperature controlled or not necessarily in healthcare? Do you see that as

a way of inspiring other industries to rethink their supply chain? I think so. I think the whole idea, if you go back like 10, 15 years, people viewed supply chain as a cost center, right? It was about how do you get product from point A to B, and some of them were thinking of manufacturing, some of them just think of logistics. If you now think of, you can create and unlock a whole new value from...

What you do from the time you're finishing the product to getting it all the way to the customer, you can create new value. And a lot of that value can be through, could be for healthcare, it could be patient safety, could be compliance. But for other industries, including healthcare, it could be hugely different customer satisfaction.

the ability to make things simpler, make things more. So you're creating a whole new value that is beyond the traditional product innovation that people are thinking. When people think of innovation, they think, okay, let's go and create a new product or a new drug. There's a whole new innovation bucket here, which is around customer satisfaction, around compliance, patient safety, that's associated with the value chain that you can go after.

How do you get people to think that way? How do you get, they have to blend an awareness of the possibilities of the technology with these use cases that you mentioned before. How are you getting internal people savvy enough to recognize these possibilities? It's a great question because typically most companies, they will have some idea of technology, but technology is evolving so fast that you may not know what else is available.

So when you create the use cases and you make sure they're compelling, then the next thing we've done is we've talked to a lot of tech partners. And then we bring in people who are experts in different areas and we have discussions and say, okay, and you start seeing some themes and trends emerge, right? And you start getting some idea, okay, this is how we can connect this between the technology and the use cases.

Ranjit, you mentioned an ecosystem of tech possibilities and partners, and I have to imagine that that also requires an ability to distinguish the pros and cons of a variety of technologies. What's your view on talent and getting the right talent, and what is the right talent for these kinds of applications? What we have done is we bring in three, four types of expertise together. One is domain experts who are

understand the industry space we operate in. They are the ones that are trying to understand the key customer problems, talk to the customers. And so we bring in domain experts. The second is then we bring in more from an IT perspective, because at the end of the day, when you do all this, you have to also integrate some of these technologies into your ERP systems, et cetera, because they're all, they have to be seamless because information has to connect.

And the third is that we bring in the tech partners. We've also brought in talent internally sometimes through consultants or full-time people who kind of are integrators. They may not be domain experts completely, but they also understand technology and they form a part of the team as well and they sit in and listen. But the critical ingredient is leadership. So let me explain that. As we have these meetings, about three or four of my senior leaders come

We sit in and listen in every session. It could be a use case session where we are talking about the customer facing use cases, the issue, the challenges from a customer perspective. There could be another session where we are talking to a tech partner. And as we do this, what happens is there's a learning that happens as you go along.

And it's very important. I cannot underscore how important this learning is for leadership to be able to. They may not, the leadership may not know. For instance, I'm not an AI expert, but I need to know how will it work? What are the limitations? How will it connect to the use case? I need to understand that. At the same time, I may not be a complete domain expert, but I need to be able to understand the use case well. So

The center of all this is there should be a core group of leaders who are constantly connecting the dots and taking help from the extended team of experts. That's very, very important. The integration of

that team into a cohesive, single-purpose-focused team has to happen through real hands-on leadership involvement. And you can't just delegate that, you're saying, to just anybody. You're saying like senior leaders and mid-managers have to be intimately involved in the fate of these kinds of opportunities. Exactly. I mean, if you think of this as a whole new

avenue for innovation, value creation, value capture. You have to have senior leadership involved. This is also a learning process for senior leadership. And what also it helps is senior leadership can make resources available. You start seeing the structural impact of this on the organization, what kind of talent you want to bring in. It's not just another activity. You have to think of this as how does this transform

the way we operate, the way we, our business model, the way our customer satisfaction approaches and everything. It reminds me of the Hamid Shah at 1-800-Flowers. I mean, there was a big theme there about democratizing and getting more and more people aware of this technology so that they could apply them as well. And again, that's another theme that seems to be coming out in many of the people we talk to.

So, Ruchi, what are you excited about? What's next? What's the cool new thing that cold chain is going to do or that you think is exciting in one way or the other, technology-wise or process-wise? What's fun and exciting? I think this whole ability to provide solutions that deliver assurance. I like calling it from plant to patient or recipient, right? The plant being the manufacturer of a life science company all the way to

the recipient or the patient. But that's, to me, it's going to be pretty exciting and a whole new way to improve patient safety, satisfaction. And I think as we do this, we'll come across new ideas as well of what we can do. As an example, I'll give you some of the things we're thinking of is we can do design optimization. We can come up with products that make more sense

And we can also give recommendations on which product to use, depending on what drug or pharmaceutical, which route, which time of the year. I mean, you can see this go in many directions where you are using data and technology to create a whole new set of value.

Ranjit, before you joining the company and bringing some of these new technologies in, there was a way of working that the company had. I'm interested in what your views are or what, I guess, transformations or changes has happened in KPIs and metrics and what's important. I mean, you talked about in quality assurance itself and customer satisfaction itself being important.

new values and new measures that are being introduced. So has there been new KPIs or new metrics that have been introduced into the company to look at the efficacy of what's happening? I guess one of the things that we have been talking about over the last one year is this whole concept of how do you provide comprehensive solutions that create value as opposed to a product?

So that's the discussion we have been having. If you tie that in with what are the big needs out there, what technologies exist to address the needs, and how do we then provide comprehensive solutions? And when you talk about comprehensive solutions, you get into product, but then you quickly go beyond product to services to data.

to different types of customer satisfaction approaches. I call it like you go from one dimension of innovation, which is product focus to multiple dimension of innovation that you could do. CCT as a company has got very talented people, very strong roots in its thermal design capabilities, testing validation.

I think the last 12 months, our conversation has been more around how do you translate all those core capabilities that this company has built over more than 50 years into providing these comprehensive solutions that create true value. Yeah, that comprehensive word is really coming out strong both here and

And, Shervin, I feel like we're seeing a theme of that with so many people who are doing well with these technologies are finding ways that they're working more holistically. Yes, they can use them to pick up on one part of the organization and improve one part of the organization, but lots of folks are starting to think about the big picture and how all these connect. And you hear trite things about removing silos and stuff, but starting to see some legs to that perhaps.

That's very true. Ranjit, what's been your biggest positive surprise as you've brought these changes in and you've brought in more advanced technologies to address these opportunities? I think the team has embraced it. I really mean that. I think we have an extremely talented, passionate team at CCT, a lot of domain experts, and

I get a lot of energy from them and they can help me connect the dots and make this bigger. We also had tremendous support from our customers and other partners. So that's kind of the approach we are taking in. I think to me, I really give a lot of credit to our team here, which has been stellar.

Ranjit, great talking with you. Your work at Coldchain really shows how pervasive artificial intelligence can be. So the connection between Coldchain and AI isn't obvious at first, but it makes complete sense in retrospect. You're not myopically moving material from one place to another, but you're creating a cohesive, holistic process. Thank you for taking the time to talk with us. We've really enjoyed it. Likewise, Sam, and thanks to both you and Sherwin. It was a great, great discussion. I enjoyed it as well.

On our next episode, our guest is Jerry Martin Flickinger, former executive vice president and chief technology officer of Starbucks. Please join us. Thanks for listening to Me, Myself, and AI. We believe, like you, that the conversation about AI implementation doesn't start and stop with this podcast. That's why we've created a group on LinkedIn specifically for listeners like you. It's called AI for Leaders. And if you join us, you can chat with show creators and hosts, ask your own questions, share your insights,

and gain access to valuable resources about AI implementation from MIT SMR and BCG, you can access it by visiting mitsmr.com forward slash AI for Leaders. We'll put that link in the show notes and we hope to see you there.