An in-depth interview with one of India's hottest startups, talking everything from robotics, capital raising and the virtues of having long-standing colleagues.
Startup Addverb Technologies builds robotics and automation solutions for factories and warehouses and can name Amazon, Flipkart, and Coca-Cola among its clients.
In this interview with Future Food Finance, co-founder and CEO Sangeet Kumar gives us a full overview of the India-headquartered business.
He tells us the business which is expected to hit $100m in revenues this year has earmarked revenues of $10bn in the next five years. He also talks about the challenges of scaling up the business, acquiring top talent and capital raising amongst other things.
Paul Cuatrecasas: Hi Sangeet, so how would you describe Addverb Technologies? Perhaps in a way we would not read on the website?
Sangeet Kumar: I was working at Asian Paints, India’s largest and one of the global giants in decorative paints. At Asian Paints, I was part of the Projects team and we set up some of the largest and highly automated paint factories you would find in the whole world. We deployed some of the most advanced technologies and developed some of the most intelligent software.
Then in 2016, we decided to start a venture in India into Industry 4.0. I was learning many new things about this space, and I thought this is something which as a country we should not miss. I felt that there should be companies that can help Indian companies grow and develop these technologies. That was the initial thought process. Then we said "OK, what form should we be? And in which industry we should be working on?
"As a user, and then while exploring this opportunity, what we found is that there were warehouse automation players who used conventional technology. So, shuttles, mother-child, ASRS, WCS, and WMS etc.
And then there were emerging companies who believed in flexible automation, so companies like 6 River Systems, GreyOrange, Geek+, Quicktron etc.
But there was not a single company that believed in both these technologies, and which believed in a hybrid of these technologies. We knew that the customer’s problem needed both technologies to find solutions. So, we decided to be that company, and that is how we started and that is what has kept us going.
So today we have solutions that fit into conventional automation, which are pallet shuttles, mother-child shuttles, ASRS and WMS etc., and then we also have very powerful WMS and all the different kinds of robots, whether it is Pick- AMR or Veloce, which I call ‘intelligent Eva’ or sorting robots, which are a very new thing, where you have 400-500 such robots working together to sort 18,000 to 24,000 items per hour.
This is something that we have achieved in the past, and this is what we are looking at. So, how we can use best of both worlds to solve the customers’ needs?
Today 40 per cent to 60 per cent of our revenue is grocery, e-grocery, general merchandising, and e-commerce. Then we have significant presence in FMCG, fashion and lifestyle, pharma and other segments. Micro-fulfillment is something which is really solving the problem of last mile delivery, the cost associated with it, and the ability to use the cubic space in the middle of cities.
This is only possible through automation, and we have the necessary products. So, we brought them together, and we tailor our solutions to suit the requirement of e-grocery, and that is how we launched our micro-fulfillment services.
PC: So, if we start with the robotics, presumably there is a platform technology that is modular so you can create different solutions around the common platform or architecture? Isn't that quite a competitive space now?
There are quite a few other robotics companies out there in terms of warehouse technology, logistics, in-store robotics and alike. What is the advantage of Addverb? Why do your customers choose Addverb for different robotic solutions?
SK: One of our biggest edges that differentiates Addverb from most of the robotics suppliers today is the range of robots that we have. We have an extensive range of robots with payload capacity of 200 kg to 1500 kg. We can carry a full pallet with up to 1.2 tonnes capacity, fully autonomously - we have LIDAR data, and we take data from the camera and fuse the data to make sense of the environment.
Our Dynamo range of products are computing heavy, and we are using cutting edge technology that is evolving fast. We are seeing the same kind of technology but at a more controlled level, and that is why the cost is going down and the software is becoming more and more complex.
These are fully autonomous robots. They do not need anything. We have shipped our products to customers without any training. In half an hour these customers were able to set these robots up for their warehouses and hospitals etc.
Then we have a second range of robots which are on grid-based navigation which use QR code markers on the floor. We use these but only for sorting applications. If you look in the market, we have the biggest range; we have table-top robotic sorters, we have ground-based robotic sorters, we have robotic sorters starting from 6 kg to 100 kg.
Today, no conventional software can do more than 40-50 kg, we go up to 100 kg. Also, the sizes of the parcels that we can handle, there is no company in the world which can give that kind of flexibility to the end customer.
The third kind of robot is a robotic shuttle, which we call Veloce, is a robot with the conventional automation feature of carton shuttle. It is intelligent in a sense that it does not bring the whole rack to the picker, but it brings only those crates which are required for the picker to be picked.
For all these robots we have a common platform, and they are all compatible with any other fleet management system as well as ours, that way we have made it extremely open.
The other thing which we have done in our fleet management system is irrespective of what robot and whose robot, the conventional automation can be handled so that task management does not look at whether this task must go to a robot or a conventional shuttle system. That is how we have designed our fleet management system.
So, these are some things which make us unique; the range of products, the interoperability, the intelligent layer of fleet management, the ability to update these robots over the air, and the robots learning from each other because they are pushing the data to the cloud for other robots to learn. This entire field of AI, ML, DL is continuously evolving, so no one can claim mastery over them. That is how we see ourselves in this space.
PC: What about on the micro-fulfillment center space because that I would imagine is going to be growing in demand, if you or others can deliver that given the need for speedy food and grocery and other perishable deliveries, as opposed to having to go to central warehouses and with long delivery times, is that a service that you offer or you enable others to create?
SK: We don't run micro-fulfillment centers, but we offer the equipment and the software that helps running them. So, in India, we are doing it for three companies, mainly wet and dry grocery, and one fruit and veg company. The requirement from the end customer for the fruit and veg example, is once the customer places an order, within 30 minutes the material must reach them.
Now, if it must reach the customer within 30 minutes, India is a very complex and traffic ridden country, so 20 minutes is the time taken to deliver in a 6-8km radius from that micro fulfillment center.
That means, in 10 minutes the material must move from order to dispatch. The rider/driver cannot take one order, it must be at least three or four orders. So, we have designed our micro-fulfillment centers to solve those issues and accelerate the process.
This company does not do a huge number of orders, around 1,000 orders per day, but for a very large chain, the micro-fulfillment centers that we design can do 7,000 orders a day, unlike in US where it would be 7,000 orders a week.
In India, because the consumption is very high, especially in cities, these micro-fulfillment centers with a 6-8km radius, have to do 6,000-10,000 orders. That is what makes us different from the micro-fulfillment centers which are being offered elsewhere in the world.
PC: Switching gears a bit, it looks like you have quite a few co-founders for the company. How did that happen? How did you all come together to say, ‘let's make it happen’?
SK: So, all our co-founders used to work for Asian Paints at some point in time. In 2008 we were building automated warehouses for Asian Paints, and then a few of the founders left Asian Paints and joined other companies and were heading supply chain and manufacturing divisions.
When I decided to jump in and get into this space, I gave them all a call and they were excited about the idea. We all came together and that is how the team got formed.
Even today, out of 520 engineers that we have, there are many people in the team, at least 40-50 of them, who have known each other for the last 10 years.
When we gave them a call, they happily joined us even at pay cuts; that was the situation four or five years ago, but today things have changed. We are very competitive in terms of salaries, employee stock options etc. and the company has grown rapidly.
PC: And where are you selling geographically? Presumably you're selling in India, but where outside of India are you selling today?
SK: Today 80 per cent of our business is coming from India & 20% is coming from Southeast Asia, Australia, and Europe. In Australia & Netherlands we have 100 per cent-owned subsidiaries; in two more months, we will have a 100 per cent-owned subsidiary in the US. Over the next five years, we are aiming to get 50 per cent of our business from India and 50 per cent from outside India.
We are going to do revenue of $100 million+ this year. In the next five years we believe that we will be reaching $1 billion revenue.
As I said earlier, out of that 50 per cent will be from India and 50 per cent will come be from outside India.
PC: Does outside India include selling in the US and in Europe?
SK: Yes. We already have a 100 per cent subsidiary in the Netherlands. We have recruited our head of sales there, and we are setting up a project office as well as the services office. In the US, we are coming in a big way, one of our co-founders is moving to the US along with a good team soon.
PC: So, the question on that growth is channel versus direct sales. Are you doing mainly direct selling still today or do you have channel partners that are signing up? How are you selling today?
SK: In India it is direct sales. Outside India, it is a mix of direct sales and channel. So, those countries where we are present, we do it directly, and those countries where we are not present, we do it through channel sales. In the US we are still creating our strategy. We are talking to certain partners, talking amongst ourselves, discussing with the board, and we will decide in few months from now.
PC: If you look out over the next two or three years, where do you think the Addverb solutions could be deployed? We could just take grocery delivery or delivery in general. You may have seen goPuff growing rapidly in the US and then in grocery there's Getir and Gorillas.
All these startups now that are very aggressively growing. They must move fast and the more automated they can be the better. Do you have views on how you see all of that developing?
SK: I feel for grocery, micro-fulfillment centers and dark stores will be an answer, due to the pandemic and convenience factors. Yes, there will be a few customers who like to touch the items before they buy, but most of the customers will move towards e-grocery, and they will want these materials to be delivered within 10-30 minutes.
So, we cannot say same day delivery is something that is revolutionary today, it has to be within half an hour, within one hour in the city limit etc. So, that is one way that grocery is going.
The other thing which will be critical, and which is important for these micro-fulfillment centers, is to see what SKUs the companies need to keep.
Today the grocery SKU or assortment itself is around 6,000-10,000 per store, but people are wanting about 20,000-40,000, but if you have this large grocery assortment and you want to give every option to all your customers, to make that practically possible, we need to delve deep into the intelligence of what sells and what are the local assortment preferences.
Every locality is different. So, what to offer and what to keep in this assortment is something which will be very data driven and those people who have access to this data will be setting up the best micro-fulfillment centers or dark stores in the country or anywhere in the world.
In the fashion and lifestyle segment we are talking about 2 million SKUs in the warehouses today, with each SKU having a depth of three units or five units. That itself is going to change.
One Chinese giant in fast fashion and lifestyle are introducing 2,000 to 5,000 SKUs every day. 2 million SKUs, which itself is a challenge, will become 20 million, because you are introducing tens of thousands more every day.
There are increasing numbers of social influencers introducing those SKUs, wearing those SKUs, and then you are launching them on the website also. So, it's going to change fashion and lifestyle as well.
These are some things which I think are going to happen, and the warehouses of the future have to be designed to take care of these things.
PC: And I presume that you are not too concerned about competition coming from China or other advanced competition in the west, like what Ocado may be doing in robotics because the market is so big that it's a rising tide, so competition isn't presumably a major concern of yours at the moment?
SK: No, not exactly. As you rightly said, the market is very large and there will be market for Chinese robots, there will be market for robots coming from Ocado etc. But I firmly believe that there will be consolidation.
The industry cannot run the way it is running today. You have 5,000 robotics companies today? Previously there were only 6-8 companies making six axis robots, you can name them: ABB Robotics, KUKA Robotics, FANUC, Fuji Robotics, Yasakawa, Kawasaki.
In mobile robots I think every day there is a new company coming up. There will be innovative companies coming from universities and building separate robots for tasks, and that will become a hit.
But for a full-fledged robotic player, it is not only the ability to produce 50-100 robots, but when you start producing robots like a car on an assembly line, it must be able to scale fast. You must sell 100,000 robots or 200,000 robots.
Every minute one robot has to come out of the factory. So, the design of the robot has to change, the design of the factory which produces such a robot has to change, and the software has to be extremely flexible and modular. Any company which can do all these three things will be the winner and they will consolidate the industry.
You will have 50-60 companies coming together or one company taking over the center by acquiring, and that is what we are trying to do, where we are changing the design of the robot.
We build everything in the robot. We build the mechanical, the electrical, the compute system, the embedded system, the vision system etc. So, it is not only the software or the hardware and taking it from someone else or from China. That is not going to work.
If you are a car company, your manufacturing needs to be a strength, whether it is Tesla or anyone else. Similarly, if you are a robot company, your software has to be a strength, your manufacturing has to be a strength. You cannot do 50 robots all with different applications and claim to be a global robotic company, that is going to change.
PC: Can you work with a company to design almost an end-to-end system for them today? Let's take a company that is delivering food or their supplier, either of perishables or even non-perishables, to supermarkets across the country across a region and they felt they were behind the times.
Let's say they were not able to keep up with the Walmarts or Amazons or Ocados of the world and they were looking for an alternative. Maybe they did not want to be wedded to one technology, but they would be open to it.
If you take a technology agnostic view, you are trying to solve their problem, could you help that type of company if it were a large enough opportunity to say "look, we can help you redesign the way that you are doing all of your logistics today? So, we can automate almost everything you're doing using robotic solutions".
SK: Yes, and that is that is the reason why I say that we will be a billion-dollar company in the next five years. We are working with a $70 billion company, which was only into offline retail. They saw the challenges coming from Amazon and Walmart, and they felt the answer to this challenge was omni channel retail.
We are working with them very closely. We have designed a mammoth warehouse for them, and they want to have the widest reach ever by a single company. We are working on the design of central fulfillment center, on the micro-fulfillment center, and on the sortation centers.
PC: What would you say is Addverb's biggest challenge going forward? What is the thing that you see is very important, that could be a problem or an obstacle?
SK: So, one of the challenges is how to scale manufacturing, and how do we produce robots per minute? We have recently hired a chief design officer who has huge experience in how the product must be designed so that it can be manufactured at scale, with minimum effort, and with our own robots contributing to building new robots, so, building a completely flexible factory. That is something which is a challenge.
I am sitting in our factory today. We are looking at building the world's largest factory, a factory that will produce quality robots at the minimum price possible. All our products have to change. What we did in the last five years is alarming, but it has to change for being manufacturable at that scale. So that is one challenge.
The second challenge today for us is getting good talent in this space. That is why we are thinking of having R&D centers outside India. We will very soon have R&D Centers in Europe and the US. It is not only that these countries become a market for us, but they also give us the necessary talent for us to build the best robots. So, these are two challenges that we think we have to solve in the next one or two years.
PC: Okay, interesting. And capital is not a challenge? The news is you are raising capital, you are close to that, but as you grow you don't think capital is a constraint as long as it's available in the market?
SK: So, capital to date has not been a problem for us and whenever we needed, we have raised. Right from the start, one thing we were very clear on was that we will not be a loss-making company. The business has to generate cash and the business has to be profitable, then only we can survive, otherwise I will get into a stage wherein I will have to take decisions which are not the best for our people, the people who have trusted us when we were very small.
So, we have been profitable, except a year when COVID hit us hard in terms of our delivery of our equipment, but I do not think there is a problem of capital with our new investor coming into the picture.
PC: On your first challenge about building the large factories, making them flexible and dynamic and changing with the times, do you see a time when the robots across the range could be manufactured locally, almost on site, with 3D printing, for example, as 3D printing itself starts to evolve more rapidly and can accommodate more complexity?
Do you see that as a possibility of evolving from the more centralized factory site that's complex to something that is more distributed? Locally, quicker, flexible?
SK: We believe in distributed manufacturing. When I say distributed there is another concept beyond micro-fulfillment centers. This concept is micro factories. You may have heard of Arrival doing it in the UK etc.
So, our new factory will be huge, but there will be several micro factories inside. Once we have established ourselves, we will take those models of micro factories to our customers, because we believe that the manufacturing has to happen near the customer, and the answer to that is the micro factory. You don't need to invest so much money and these must be extremely flexible.
So, distributed manufacturing and micro factories, yes. As far as 3D printing is concerned, I still believe that it has a long way to go. We use carbon fiber 3D printed parts for their strength, and we are using it to build our own cohort, but that is still limited to prototyping or manufacturing 20-50 robots a year, but that is not what we are talking about. We have to produce 50,000-100,000 robots a year.
So, if that has to happen, I think 3D printing has to go a long way. I know Boeing is using it in airplanes and GE is using it, so we can use it, but it should also make commercial sense for us.
PC: Sangeet, I could talk with you with all day, but I know it’s getting late for you so I’ll let you go. Thank you so much for your time.