Introduction To Nidhi
Tom Raftery: [00:00:23] Good morning, good afternoon or good evening wherever you are in the world. This is the Digital Supply Chain podcast. Number one podcast focusing on the digitisation of supply chain. And I'm your host. Global Vice President of SAP, Tom Raftery. [00:00:36]
Tom Raftery: [00:00:38] Hi, everyone. Welcome to the Digital Supply Chain podcast. My name is Tom Raftery with SAP and with me on the show today. I have Nidhi. Nidhi, would you like to introduce yourself? [00:00:47]
Nidhi Gupta: [00:00:48] Hi, Tom. Really glad to be here. My name is Nidhi. I'm the co-founder and CEO of Portcast [we are] based in Singapore and are a predictive analytics company, working with logistics companies and shippers.
We predict how ocean cargo moves across the world. So we predict the demand and the arrival times of any container in each container vessel around the world using machine learning and various advanced data sets. Really glad to be here. [00:01:14]
Introduction to Portcast
Tom Raftery: [00:01:14] Thank you. Glad to have you. Before we turned on the recorder here, we had a brief conversation and you mentioned that you are a Start-Up. You've been in operation for two years. [00:01:23]
Nidhi Gupta: [00:01:24] That's right. Yeah. So we're based in Singapore. We're a VC backed Start-Up. We are a team of data scientists, software developers and industry professionals who come from the logistics space would be working with companies globally. I think technology has made that possible that we can. So technology from Singapore. But for the world. [00:01:45]
Tom Raftery: [00:01:46] Technology's made a lot of things possible. We also talked before we turn on the recorder about COVID and lots of people are now working from home where that wasn't possible before. [00:01:53]
Nidhi Gupta: [00:01:54] Yeah, absolutely. [00:01:55]
Why Is Predicting Container ETA Important?
Tom Raftery: [00:01:56] So tell me. I know very little about logistics, so let's just get that out there. So explain to me like I'm five. Why would I want to know the departure or arrival times of containers? [00:02:10]
Nidhi Gupta: [00:02:12] Yeah. So, you know, interestingly, we don't see shipping as we see trucks, but 90 percent of the things that you see around you, the like your sofa or the microphone or the headset, everything's been on the ship at least once in its lifetime. Trade essentially moves on a ship. So shipping is an integral part of getting the good stuff we have to ourselves.
So what happens in shipping really makes the costs, the efficiency and the lead times of how we're getting our goods really meaningful. And on top of that, it is one of the major components of carbon emissions.
So, you know, with the entire climate change conversations, it becomes really important that we start talking about how we can make shipping more efficient, which not only impacts you and me as a consumer, but it also impacts the economy and the environment. [00:03:17]
How Portcast Provides Supply Chain Visibility
Tom Raftery: [00:03:17] And how? Again, like I'm five. How how does your solution make that possible? [00:03:24]
Nidhi Gupta: [00:03:26] So if you think about logistics, it's a very fragmented market. There are multiple companies that basically come together in order to bring a particular good from China to Spain. There are multiple trucking companies. There are multiple shipping companies, multiple ships. So essentially it's of a fragmented and disorganized market.
All of these handovers mean that there's no single source of information around the end to end visibility or traceability of the cargo. And on top of that, there are so many changes that can happen.
Planned or unplanned ships can change the routes, shipping companies can change schedules, but they could be due to unforeseen events like the blasts in Beirut or the US-China trade disputes.
All of these disruptions make it even harder to get real visibility and predictability. Daddo in terms of the supply chain. And that's where we come in. We basically use vested assets and machine learning to enable one single solution. But you can see the visibility of how your containers are moving, how your cargo is moving and when it's going to actually arrive. [00:04:37]
Why Data From Vessel GPS Is Not Enough
Tom Raftery: [00:04:39] OK, but don't ships have like tracking beacons the same way planes have. So you can see where a ship is at any point in time. [00:04:48]
Nidhi Gupta: [00:04:49] Yeah, absolutely. That's one of the data sets that we use. So ships are sort of regulated that they have to send satellite information. They like to do longitude of their location. But what's the next port that they're going to?
When are they going to arrive at the next port? It's up to the captain of the ship to actually feed that information in terms of when is the final arrival time. They don't have to do it every minute or every hour. It depends on them how often they want to update it. So it's available with a lag or outdated or static information.
In that sense, it's not really real-time. Secondly, that information is only telling you about the position of the ship. Right now, it's not really telling you about the movement of the entire ship. So imagine, you know, getting a delivery from Amazon, it shows the G.P.S. of the truck crate down to where the truck is right now.
For example, the driver would say "I'm going to IKEA". Right. But it does not say anything about the traffic congestion that the driver is likely to see or the roads may be blocked because of construction or, you know, any other disruption that could happen. That's exactly what's happening with that G.P.S. of the ship, essentially. [00:06:06]
Who Are the Customers of Portcast?
Tom Raftery: [00:06:08] OK. So who would be a customer of Portcast? Typically. [00:06:14]
Nidhi Gupta: [00:06:16] So two kinds of customers. It would be logistics companies, so large freight forwarders who basically manage the goods movement for any manufacturer. And secondly, the manufacturers themselves so that they can provide better value-added services to their customers, and they can tell them exactly where the cargo is.
What is going to be at risk? When is it going to arrive? Is there a risk that they're predicting and so that they can manage their downstream supply chain better?
With better visibility, they can get the trucks just in time. They can communicate to the customer just in time and basically improve their customer retention and service levels. The manufacturer needs this information so that they can get better control of the supply chain for themselves, not just having to rely on the logistics company
If the cargo is delayed and there's no truck available to pick it up and the cargo has to stay at the port, the manufacturer has to bear those charges. So it makes even more sense for the manufacturer to actually get this information in advance.
The other thing is that imagine an automotive customer. And if the one single port is not available, the factory line could come to a halt. And that's an extreme cost impact for the auto company rather than, the cost of getting that predictive information and getting a notification that there's going to be a delay and then they can manage that delay much better
So those are the two kinds of customers that we deal with. [00:08:01]
What Data Sources Does Portcast Tap On?
Tom Raftery: [00:08:02] And the information sources that you use, what typically would they be apart from the beacon on the ship. [00:08:08]
Nidhi Gupta: [00:08:08] So there's definitely that geospatial information, which is the beacon on the ship of the A.I.S data. That's it. That's what it's called. But also weather disruptions, really granular information in terms of the wind speed, wave height.
What are the current cyclones and so on. The other source of information that we get directly from is port information, because ports, a lot of the time have, you know, really accurate information about when the ship is actually arriving when the container is actually getting on to the ship and so on.
We get the data directly from shipping companies as well in terms of the vessel schedule and if they're making any changes to that. And then we get external information about what's going on at the ports. Is there likely to be congestion? How many ships are waiting at the port at a particular time in the year?
So these are the multiple data sources. And then the last thing would be economic patterns about what's the capacities of different ports. What are the consumption and production patterns of different countries? And what can we see from the economy in terms of how trade is likely to change and how that impacts shipping? [00:09:21]
How Machine Learning Is Implemented at Portcast
Tom Raftery: [00:09:22] And you run that all through, you said, machine learning. [00:09:24]
Nidhi Gupta: [00:09:26] I think the good thing is that data is all available - all of this data that you're talking about is out there. The hard thing is how we can get this data to speak the same language. You know, how we can secure it in a centralised manner, in a secure manner. Keep it updated realtime and then sort of start creating a derived data SAP's on it.
So that's really the challenge. And that's the uphill task, which takes some considerable effort in terms of, you know, sort of synching all of the data together. And then it becomes easier to then use machine learning to start making sense of this because it's so many different data points that I'm talking about.
That's the machine learning can do it better than human intuition. And human intuition is what the logistics industry has primarily been using from spreadsheets to emails and phone calls. So that's why machine learning starts making things a lot more digestible and enables actions. [00:10:23]
Tom Raftery: [00:10:24] OK. What kind? Can you speak to successful use cases or success rates or anything like that? [00:10:31]
Nidhi Gupta: [00:10:33] So if you look at baselines of how companies have been predicting demand and arrival times, it's been about 70 to 80 percent accurate. What we've been able to achieve in terms of machine learning has been almost 90 percent accuracy.
And that is significant because if you look at any particular ship if you try to get the utilization of any particular ship up by even one percent, that is millions of dollars for a single ship.
If we tried to predict with high accuracy five days ahead, the manufacturer has so much more time to plan their supply chain better and that saves them time, lesser safety stock, and it saves them the cost of expediting through airfreight.
So it becomes really valuable, even getting single percentage improvements. And I think we've been able to see tremendous improvement in the accuracy levels. [00:11:31]
Impact of Supply Chain Visibility on the Logistics Industry
Tom Raftery: [00:11:32] How has it impacted the logistics industry? [00:11:35]
Nidhi Gupta: [00:11:37] You know, I only see it as a positive impact in the longer run. I think COVID has kind of increased the urgency for logistics to adopt technology and to digitize much faster. We've seen during COVID how important logistics is until now.
Initially, it was mostly seen as a cost factor in companies, and now people are realizing that the most important thing is how we're going to get those goods to our home because we're not going out there and getting those goods.
It all comes down to supply chains being a strategic enabler for getting that competitive advantage. And so when logistics starts becoming important, companies know that they need to then reduce on-time delivery issues.
So they need to reduce the inefficiencies. They need to become more technology-savvy so that any lead times that are lost, any leakages in the supply chain are gotten rid of. And that's where I think we've seen a lot more interest from companies during the last few months.
And I think we're kind of at an inflection point. So the next two to five years will see large scale tech adoption in the logistics industry. [00:13:14]
Examples of How External Disruptions Lead To Shipping Delays
Tom Raftery: [00:13:24] And have you seen any impacts directly on shipping lines of COVID-19, delays, impacts on manufacturers, anything like that? [00:13:33]
Nidhi Gupta: [00:13:33] Yeah, absolutely. I mean, just today I was talking about a few examples of what technology is able to see. So, you know, there was a recent cyclone near the Panama Canal. And we saw ships that were taking a detour because of that cyclone. The shipping company was saying exactly that the ship is actually going to be on schedule and it's going to come at the time that they had promised.
But we knew that this ship is actually not going to be on time. It's going to be delayed by at least a day, at least 24 hours, just to the next port. And then there's a knock-on effect of all the ports beyond that. And that's really what we're able to say.
If they have gotten this information in advance, they could have adapted their supply chain.
So that was one example. The other example is the recent Beirut blasts. We saw ships that were going to go to Beirut on the day of the blast. These are ships that arrived on the day of the blast and were supposed to leave the next day.
We had predicted that these ships are s at least going to stay at the port for three days because of all the congestion that we were seeing around the port. And eventually, the ship actually stayed for days. So we were really accurate.
And we were able to predict that much in advance so that those shipping, the logistics companies and the ships can get that information. So I think what we're talking about becomes really tangible with these kinds of examples. [00:15:12]
The Entire Logistics Journey When You Order Something Online
Tom Raftery: [00:15:13] Let's change tack for a second, because, I mean, this whole industry is very niche. I got to think I know, as I said at the start, very little about it. Well, let's say I go on to Amazon and I order a phone or I order a piece of my computer or whatever, and it arrives the next day.
But you mentioned at the start that everything around me has been in a ship at some stage. Ships coming from China to Europe probably take maybe three or four weeks, I don't know, depending on your point earlier and how many stops they make en route.
So how does all that work? That I order something from Amazon and it arrives the next day? Typically the goods are from China, loaded on a ship, transported for 30 days and then stored in a warehouse. It must be something that I'm guessing. [00:16:05]
Nidhi Gupta: [00:16:06] Yes, absolutely. Yeah, you're right there. So basically, there would have been a factory in China, which is an OEM, let's say, you know. Based in China. The goods would have been filled in the factory in a container. The container goes on the road to the nearest port in China, let's say Shanghai.
And then from Shanghai, it needs to reach Spain. So it takes about two to three weeks, about 20 days to get to Spain. And then at the port in Spain, the nearest port to your hometown, it then needs to get on to a truck or maybe a train. And then it needs to get delivered to a particular warehouse of Amazon.
It gets there and gets stored there. And Amazon plans that inventory by forecasting the demand. When you're more likely to buy the phone, based on that. And it sort of brings in the shipments from China based on the forecast. So it stocks them in the warehouse in Spain. And when you order. It just needs to deliver with, you know, in the parts to your home. And that's what it can do in a day's time. But what Amazon has to do on their side is to plan that inventory.
So they have to focus that demand based on the consumption patterns they are seeing. They have to plan in advance how long it's going to take for the lead time between China and Spain in terms of the ships that are available as well as the disruptions that might happen.
And then they have to place an order much in advance before that with the factory in China because it needs to get from the factory to the port in China as well.
So all of that is happening behind the scenes. Imagine if Amazon does this all day, and yet there is a cyclone and the ship doesn't arrive on time and it misses that window. You may not get that phone in time and then you're not very happy with Amazon. So those are the things that are happening behind the scenes. [00:18:12]
Why Did Nidhi Start Portcast?
Tom Raftery: [00:18:13] OK. You're a two-year-old company. What made you get into this? Because like I said earlier, it's a very niche topic. [00:18:19]
Nidhi Gupta: [00:18:20] Yeah, I know. So my background has been in logistics. I've worked for Deutsche Post DHL for a decade in Singapore. But my role was all across Asia Pacific. I did management consulting, business development and strategy for them. So I was always dealing with customers on issues of C-suite executives and customers in terms of pain points that they have with their supply chain.
And it always came down to inefficiencies and lots of costs. And we were using Excel sheets. We were using emails. We were following up manually. Not many companies were digitised at that time.
But Amazon was really gaining prominence because we were talking about Amazon. If you just look at warehousing, we saw Amazon rapidly innovating. They had robots throughout their warehouses.
In contrast, logistics companies take time to experiment. And before they can make it a global phenomenon, they have to really think about the pros and cons and the business opportunities and viability of it before it becomes something scalable.
But there is a real advantage to implementing technology, which is really visible in terms of how technology impacts efficiency, costs, time, and eventually customer satisfaction. And therefore, I was really sort of excited by the role that technology plays in this space.
And shipping was what happened because, you know, as I said, there's a lot of inefficiency in the space, which I see as an opportunity to improve. But eventually, the idea is that what we're doing has a role to play in the entire supply chain and we could expand throughout the supply chain eventually. [00:20:11]
Making Shipping as Efficient as the Airlines Industry
Tom Raftery: [00:20:13] Superb. We're coming up on the 20-minute mark now, which is typically when I start to wrap up the podcast. Is there any question I've not asked that you wish I had asked or is there any topic that we've not mentioned that you think it's important for people to be aware of? [00:20:27]
Nidhi Gupta: [00:20:28] I think it might be surprising to people. What if this is so important? Why are companies not doing this already? So, you know, shipping and logistics has been quite an opaque industry. You know, if you think about the passenger airlines, when you buy an airline ticket, it's almost an instant process. You can get multiple options. You can book a ticket. You can get the price dynamically.
And you know exactly when the flight is going to arrive or if there's a delay and it's taking off. And it's all in minutes. This is not where the passenger airline industry was two decades back. It was still with agencies where you had to book tickets. It was still pretty manual.
But in the last two decades, they have strongly changed how the industry behaves and they've kind of improve efficiency and made everything digital. And that's what we think of as normal now.
Shipping is still a two-decades-old space. And, you know, not just the entire logistics industry, but we're seeing the last mile with Amazon and with trucking. We're seeing that change coming. The upstream, like air freight and ocean freight, still has a way to go.
Companies want to make that change, but they are still not at the stage where they should be. So I think there's real value in making that happen. It's a lot to do with the change in management mindset. The industry has been talking about it for far too long. And I think and I hope that it kind of makes that urgency really, really stronger. [00:21:59]
How You Can Reach Out to Us
Tom Raftery: [00:22:00] OK, super, super Nidhi. That's been really fascinating. If people want to know more about Nidhi or about portcast or about shipping in general. Where would you have me direct them? [00:22:12]
Nidhi Gupta: [00:22:13] I'd be most happy to connect with them on LinkedIn or they can reach out to me at firstname.lastname@example.org [00:22:28]
Tom Raftery: [00:22:29] Oh, fantastic Nidhi. That's been great. Thanks again for coming on the show. [00:22:32]
Nidhi Gupta: [00:22:34] Thank you, Tom. [00:22:34]