Blogs

AsiaIndustrial NetNews: On May 9, local time in the United States, at the XPo just heldAt the nENTIAL 2017 conference, Intel CEO Krzanich ordered a bottle of drink using his mobile phone, and then a deliveryrobotWalking slowly onto the stage, Korzaiqi used the password on his mobile phone to open the box on the robot and took out the drink.

If you look closely, the rear two wheels of this robot look like a balance car that is common in our daily life. In fact, the robot is not unrelated to the balance car. This is a robot under Segway & Ninebot. Loomo Go, a courier robot jointly released by Segway Robotics and Intel.

Loomo Go delivery robot

Loomo Go is a delivery robot with self-driving capabilities. It is actually developed based on the Segway robot Loomo. It is mainly used to improve the efficiency of last-mile delivery and replace the repetitive and time-consuming walking of deliverymen with Robotic technology. labor.

Different from many current concept land delivery Robots, Loomo Go adopts a dual-purpose design of automatic and manual driving. In this regard, Pu Li, vice president of Segway & Ninebot, told Lei Feng. The dual purpose is to allow the robot not only to save labor, but also to work with couriers to flexibly handle complex situations in real scenarios.

In many application scenarios, the combination of manual and automatic is very necessary. In some cases, some links are considered to be more efficient, depending on the application scenario. For example, when delivering food in a shopping mall, because the environment in the shopping mall is relatively simple, no one is needed in this case. But if the environment is complex or the tasks to be performed are complex, manpower is needed to solve it.

Because our platform itself is a new universal robot product that is programmable, movable and extensible, we did not consider designing it as a device for a certain scene when designing Loomo Go. A machine that can both carry people and operate automatically may be applicable to more scenarios. In industries such as distribution, logistics, security patrols, and escort, there are still a large number of objects or people that rely on the movement of workers on foot. What Loomo Go offers is the ability to gradually replace the labor of reciprocating walking in a walking environment.

1 2 Next page > page

Robots that move reliably indoors and out

According to reports, this Loomo Go delivery robot based on vision and radio-assisted navigation can achieve centimeter-level positioning accuracy indoors and outdoors, and can achieve seamless connection between indoor and outdoor actions in an unmanned delivery environment. According to Segway Robotics, similar Its visual navigation technology can be found in self-driving cars, Google’s AR phone Tango, and the iPhone 8 predicted by various senior analysts.

Compared with self-driving cars and handheld devices on the highway, Loomo Go has been deeply optimized in the application scenarios of mobile robots in walking environments, giving it higher accuracy and good stability in mixed indoor and outdoor scenarios.

Pu Li told Lei Feng.com that “New Intelligent Manufacturing”, in places where high-rise buildings are particularly dense, GPS signals will be blocked, and the vision-based multi-sensor fusion solution can make the robot more stable. Previously, when Segway Robotics tested indoors, the robot moved in a relatively large circle and walked for several hours without error.

Pu Li said that there will be a hackathon at MIT this weekend, when Loomo Go will deliver drinks to the players on the spot, which also happens to be a practical test.

Cooperation with Intel

When Lei Feng.com asked about the roles of Intel and Segway Robotics in the development of Loomo Go, Pu Li said that, like the previous PC, smartphone and other industries, robotics technology must be applied to the ground. It is necessary to have upstream key sensing and computing devices, midstream robot host integration and integration, and downstream vertical application scenarios.

What Segway Robotics does is to focus on midstream integration and software and hardware integration, gather cutting-edge and excellent upstream supply chain resources, and provide rapid hardware prototypes, mobile navigation subsystem services, and easy-to-use algorithms for the development of downstream robot applications. middleware.

Therefore, “integration” and “platform” are the strengths of Segway Robotics, and in the Loomo Go machine, Intel provided it with important components or solutions such as RealSence. With Loomo’s foundation in place, the rate at which bots can unlock new skills is also accelerated.

As I said before, Loomo Go is an extended version of the previous Loomo robot. Due to the foundation of Loomo, the development cycle will not be particularly long. At the CES in January this year, German BMW only took 35 days to build a robot based on Loomo II. This time, he developed a “self-driving car parking assistant” that worked 8 hours a day for 3 consecutive days in a Las Vegas parking lot to receive audiences participating in the autonomous driving experience. The development cycle of Loomo Go this time does not exceed 2 months.

Business

The deployment cost of Loomo Go is only equivalent to 3-6 months’ salary of an ordinary courier, and its cost efficiency can be multiplied.

Although the official price of Loomo Go is not given, we can roughly calculate a range. Regarding the future of this machine, Pu Li told Lei Feng.com that “New Smart Manufacturing”. At present, the developer version has been purchased and used by customers. In Q3 of this year, it should be used by more developers. Reach a production level state. The platform’s data services will be launched in Q4.

In business, it can be divided into two parts, one is the platform, Segway Robotics will continue to play the role of a data service operator, and Loomo Go will choose the business model for sale.