回到网站

Introducing Loomo: When Mobility Meets AI - with Segway Robotics President Li Pu

· The Boss Interview,Segway,Xiaomi,Robotics,Ninebot
A photo of the Loomo robot

The Harbinger recently caught up with Li Pu, the President of Segway Robotics (division under Segway-Ninebot Group). In an earlier interview, Li introduced us to Loomo, Segway-Ninebots revolutionary robot with autonomous and self-balancing transport capabilities. In this interview, Li Pu updates us on the consumer launch of Loomo and shares his thoughts on Loomos various use cases, the B2B vs. B2C robot market, mass-market robot appeal, and future developments

Transcribed and Edited by Caroline Chen

[Editor's note: this interview has been edited and condensed for clarity

Disclosure: Adam Bao is currently consulting for Segway Robotics]

***

Loomo’s Consumer Product

Caroline: Thanks for joining us again! Why don’t you start by giving us a quick update of what you’ve been up to since our last interview?

Li Pu: We’re launching Loomo at CES in January 2018 in Las Vegas. Because CES is a trade show, all consumer electronics companies go there. Compared to big names like Samsung or Huawei or Xiaomi, Segway Robotics is still relatively small, so we want to show the world that we have created something innovative. We'll have a big booth this year debuting Loomo programmed with software that we’ve developed for consumers. This means that when consumers get Loomo, they won’t have to program Loomo themselves – Loomo will come with pre-installed capabilities for functions like entertainment, gaming, etc.

We know that Loomo’s concept is very new. Segway has been around for quite a while, so almost everyone knows what Segway is. But when people talk about robots, they have a lot of different ideas. When people talk about a vacuum cleaner robot, they have a clear concept of what it is because it's been around for a while. For products like Loomo, there are not a lot of similar products. (see CES 2018 coverage with The Verge).

So at this moment, in terms of marketing strategy, we want to advertise Loomo as a product for innovators and early adopters. Gradually, we want to build a core community where innovators and early adaptors can interact Loomo and come up with more ideas and use cases. At the same time, we'll continually update our software. Because Loomo runs on Android software, everything can be updated through OTA, similar to how you update your phone. We'll push out new functionalities every month after the product launch. The new functionalities will be based on feedback from these innovators and early adopters. It will be a long process, but we want to start from here and gradually expand the user base with the aim of entering the mass market.

We’re hoping to build up our U.S. office as well. Segway has had a U.S. presence for a long time and we've been working with our colleagues closely. But to properly compete in the robot and the self-balancing vehicle markets, we have to bring on more talent, especially on the marketing side so we can better understand the requirements and demands of the U.S. market and give that feedback back to R&D to improve our product. This goes for the European market as well.

We’re also very grateful for Intel’s support. Intel has supported us from the very beginning. To us, Intel is a giant, and we're very grateful that Brian Krzanich (Intel’s CEO) has been looking at our product.

A picture of Loomo Go

Loomo Go

Caroline: What makes Loomo stand out from its competitors?

Li Pu: I don't think there are any direct competitors because Loomo is a very unique product. It’s the only consumer robot product that can be used versatilely, and is the only one that can be used in an outdoor environment. You can use Loomo both outdoors and indoors like in this office here - well, you'll probably want a bigger room. A lot of consumer robots are very small; they can easily be moved around but they can’t function in an outdoor environment, only on flat surfaces like a table or a floor at home. They’re not designed to move on terrains like grass or a bumpy road surface, so Loomo is very unique because it can go in those areas.

I do think that the concept of robots has started to be accepted by consumers. I see many other products, like robots that can sit on your desk and communicate through speech and voice. There are also robots that you can use your cell phone to control. All of these products are very successful and it's a good sign that consumers are getting used to robots! But all of these concepts are still very new. I don't think Loomo is a product for the mass market at this moment because it will take time for people to understand what Loomo’s capabilities are, as well as its limitations. This whole industry is still in the early stages. It's going to take some time to move beyond early adopters and innovators.

Loomo's components

Loomo's components

Caroline: Why are you focusing on the consumer market with Loomo and not the enterprise market? Do you envision, at some point, servicing the B2B and B2C market with the same product?

Li Pu: We've been developing Loomo as a platform, which is also its advantage. If we developed a new product for every new application, it would take us a long time. Each of these new applications only takes us one to two months to implement because Loomo is a platform.

In terms of B2B, we’re currently playing in a B2C market. I think the B2C market will definitely be much bigger but will take a longer time to mature, because for consumers, you really need time to first let the consumers know about Loomo, then to let them develop trust in the product and become familiarized with its functionalities, limitations, ultimately unlocking the value Loomo can bring to the consumer. It'll take a longer time, and only at that point will this product reach the mass market.

For enterprise B2B applications, like security and commercial environment robots, the economic argument is very clear: Loomo will do things more efficiently and save labor costs. Right now, a lot of the work done by humans - security guards, for example, not a lot of people actually want to do this work. It's boring and repetitive. For this type of work, this is an area where robots can easily increase efficiency. They also don’t get tired of walking long distances.

So a robot’s success is measured by the economic rationale. If the price point and cost of manufacturing these products is low enough, then when the economics turn to positive return on initial investment, the market will be ready.

Caroline: Can you take a bold guess as to when you imagine that point coming?

Li Pu: I think we're already there now. The cost of our product has decreased dramatically. It is critical that Loomo is comprised of commercially-available, off-the-shelf components. Each component has been used elsewhere in consumer products in at least a million units. We chose these components deliberately because we knew that this was the only way we could get cheap, reliable components, and when we combine them all together, that's how we get an affordable platform. This has been our thinking since the beginning.

Loomo has been a consumer product since the beginning and we've always kept price affordability in mind. The quality of Loomo far exceeds what consumers expect for its price, so it's only a matter of marketing strategies and how we optimize the functionalities when we work with B2B customers. But B2B customers also have a lot more patience. They can give us feedback, help us understand the problem, and wait for a fix or upgrade. Adapting Loomo to the needs of the B2B industry will take less time than the consumer market getting used to robots.

***

Security, Delivery, and Commercial Solutions

Caroline: Can you give us a brief overview of the B2B side?

Li Pu: We've been delivering a lot more robots to our customers and developers, especially to universities and companies that have an innovation department. We’ve noticed three major use cases: security solutions, delivery solutions, and commercial solutions.

Many security companies are interested in using Loomo. One such application is they can install additional cameras on top of Loomo and let Loomo autonomously navigate in places like parking lots. I’m thinking of a U.S. partner – they’re looking at parking lots where there are a lot of license plates. In California, for example, the DMV and police will provide a list of suspicious license plates so if Loomo sees one of these plates, Loomo will alert the security department. There are also consumer applications. For example, if you parked your car but forgot where you parked it, Loomo can navigate through the parking lot and guide you to your car. This is one big application of our technology - security.

There’s also a huge demand for delivery applications. Loomo is very advanced in mobility. Loomo is by no means as fast as a Tesla, but it's adept at navigating crowded places. When you drive on the road, there are lights and rules. So all that autonomous self-driving cars need to do is follow the same set of rules. And the rules already exist. But for robots, in an environment like this office, there are lots of obstacles so robots need to constantly make adjustments to avoid people and other things. To me, even though this environment is lower speed, the increased unpredictability makes the technology much more complicated than that of autonomous cars. So we've been doing a lot of work on this. We launched Loomo Go with Intel in May 2017, and Intel CEO Brian Krzanich showcased Loomo Go at XPONENTIAL 2017, a conference focusing on autonomous applications.

Since then, we've gotten a lot of inquiries about this technology from many delivery-intensive companies. In San Francisco, robots are already on the sidewalks delivering food from restaurants to customers. We have a slightly different approach because we're in Beijing. In China, there are more than 10 million food deliveries a day. This market is growing 35% every year, which means that every two to three years, this will double. There will be a big gap between demand and delivery couriers. There's not enough human labor to fill this demand, so this is an area where robots have huge potential. And it’s not just delivering food, but also other types of goods and household items like groceries. So this is the second area we're looking at, Loomo Go.

Intel CEO Brian Krzanich introducing Loomo Go

Intel CEO Brian Krzanich introducing Loomo Go

The third area we're looking at is commercial environments. Regarding "new retail," the idea is that sales penetration reaches only 20% of its potential if you sell products online, and 80% of sales are still offline. The success of online selling comes from the fact that all the actions - every click, every purchase - is recorded, so e-commerce companies can use this data to precisely target a potential customer. But in the offline world, you don't have very precise tracking. For example, in Family Mart, you'll notice that the cashier has a huge keypad. The keypad has very rough categories: gender, age ranges like 10-20, 20-30. Cashiers have to push these buttons every purchase to collect data. It's not very precise. But now, we have AI-enabled technology to recognize faces and things like age. This is very cool – it’s the contact point of offline and online environments. So robots can do the work that the store staff is doing now, things like keeping track of customer flow and demographics, and it will be more efficient. Having a lot of cameras in stores is also a bit disconcerting for customers, and most cameras are installed on the ceiling which isn't the best angle to see faces. But robots can be situated next to the entrance of a store or move around and record this information easily. We’re confident Loomo can work well in commercial environments.

A picture of Segway-Ninebot’s product offerings

Segway-Ninebot’s product offerings (Source)

***

Rolling on into the Future

Caroline: Taking a step back, what has the reception been like since you launched Loomo? Has there been anything that surprised you?

Li Pu: Since CES 2016, a lot of developers around the world - over 3,000 people - have shared their ideas about what they want to do with Loomo. Some ideas were a bit far-fetched, inspired by sci-fi movies and TV shows. But some can achieved right now with current technology. I think current technology can provide reasonable utility to customers who will pay a reasonable price for it, and we're working on finding a match between what we can provide in terms of functionality and price and what can be done on the application side. We'll keep watching what Loomo can do. If we see any application that makes sense and creates real value, we'll look into it. We'll support our customers who have application knowledge. All told, our dream is to create a platform with mobility and computer vision.

I’ve learned a lesson recently: a lot of times people will say that the only thing that limits you is your imagination, but sometimes the thing that limits you is the imagination itself. Oftentimes, people imagine things but don't have a clear path to get there. We definitely have to have a lot of creativity and imagination, and we want to share our vision and amplify this imagination by inviting developers and community members to join us. But there also has to be a realistic path, something that can be achieved based on current technology and predicated upon a reasonable business model. It's all about striking a balance - sometimes we have to take a step back from our imagination and find a real business. You have to find a reasonable ROI so that the company – as well as the overall industry – can continue to develop.

Caroline: On the topic of innovators and early adopters, let’s talk about Double 11, or双11. Ninebot announced record-breaking sales on 双11, with sales totaling 50 million RMB, 2x those from last year. What do you think this signifies? Is there a change in consumer perception beyond just early adopters to more general interest?

Li Pu: Comparing this year’s double 11 versus last year's, we made almost twice the profit. We had launched the Xiaomi-Ninebot Mini around October or November 2016, so now it's been out for about a year. I think that most first-year buyers are primarily early adopters, but now it's starting to enter the mass market. My guess is that because information is now a lot more widely available, people are accepting novel concepts and products faster than before.

Segway-Ninebot’s Double 11 Sales

Segway-Ninebot’s Double 11 Sales (Source)

If you look at how long it took for the vacuum cleaner robot to enter the mass market, it was invented around 1998, but only in the last two to three years or so did I notice a lot of people using it in China. At Segway-Ninebot, it has only taken us two years to make a robot most Chinese people have heard about. It's very fast, but it’s also a challenge, because as a smart device hardware company, we have to be fast enough and innovative enough to get people's attention every year. We have to launch new products every year and the products have to meet the consumers' expectations. It's an opportunity but also a challenge.

Caroline: In other recent news, on October 17th, Segway-Ninebot announced a $100 million USD Series C round, with funding coming from the China Mobile Innovation Industry Fund and SDIC Innovation. We're curious what this government support signifies, in your perspective, for Segway-Ninebot’s future.

Li Pu: I think it's a very good sign for the advancement of AI-driven technologies. Two years ago, only venture capitalists were looking at this area. Now, government-backed funding is entering the space. It's a good sign that this is generating real value for society, and that the value is being recognized by a broader group of people.

I'm not in charge of financing, but support of AI and robotics initiatives is a very big portion of the investment. Another portion is support of self-balancing, short-distance transportation technology. Segway-Ninebot’s vision is that we want to move people and goods more efficiently. Mobility is very deeply rooted in Segway-Ninebot and I think this funding is primarily for a big push on this front. It's a good signal for future growth.

Caroline: Thanks for joining us again today Li Pu! Very interesting insight.

The opinions expressed in this article are Li Pu's own and do not reflect the views of Segway Robotics