Russian Marvelmind is interested in Boeing, GM, Motorola and Amazon
The founder of Marvelmind Robotics, Maxim Tretyakov, began to develop his startup as a hobby, working in parallel in one of the divisions of Nokia. And in 2016, he already received the first prize at the SLUSH startup competition in Shanghai. The company has come up with an alternative way to navigate robots indoors.
Initially, Marvelmind Robotics developed a design platform for designing robots, but in 2010 Tretyakov decided to focus on one important problem of robotics-indoor navigation. Unmanned devices, including robots, cars, drones, etc., are relatively confident in navigating in an open space using GPS and GLONASS satellite navigation systems, and indoors, in theory, you can use navigation via a Wi-Fi or Bluetooth signal. But the accuracy of positioning using Wi-Fi or Bluetooth is not enough even for the device to get into the doorway. There are also lidars that are used in self-driving cars, but they are bulky and expensive and are not suitable for home devices. Maxim Tretyakov suggested using ultrasonic sensors for navigation.
When positioning using Wi-Fi or Bluetooth, the signal strength is used to determine the location of the object (the farther the object is, the weaker the signal). Ultrasonic sensors and lidars use the time of signal transmission to a beacon, wall or other object for positioning, which allows achieving greater positioning accuracy. The signal source is located at a certain point in the room, and by analyzing the signals from many beacons, the solution calculates at which point the robot or other moving object is located. The very idea of using ultrasound was not new, but Marvelmind Robotics managed to develop a comprehensive software and hardware solution that provides the necessary data analysis speed, noise filtering quality, coordinate calculation speed, etc.The solution as a whole provides indoor positioning with an accuracy of up to 2 centimeters with a sufficient update speed to ensure smooth movement, for example in VR games.
Getting out into the world
To inform the market about its decision, Marvelmind participated in various industry exhibitions and competitions. For example, in 2016, she took the first prize at the SLUSH international startup competition in Shanghai. So far, the main type of consumers of products are companies that are studying the possibilities of using ultrasonic navigation. Every year, Marvelmind receives hundreds of orders from all over the world, while they mostly buy several sets. The minimum cost of the set is $399, depending on the customer’s needs.
Gradually, the startup attracted the attention of many major players: Amazon, Boeing, GM, Motorola, Caterpillar, Tata, Lufthansa, Porsche and others bought lots of beacons and Marvelmind software.
“Many customers order our solution as part of their larger projects, the details of which are often not disclosed, but, depending on the customer’s needs, the requirements for our solutions vary greatly, so one of our main tasks was to develop a line of solutions for a certain number of typical scenarios — for example, if our sensors are used to control unmanned loaders in a retailer’s warehouse, for navigation in a mine or for controlling drones. For drones, the most important thing is the weight of the device, and for miners, it is important to work in difficult conditions and fault tolerance, ” says Maxim Tretyakov.
Not Single Robots
In addition to robotics, Marvelmind Robotics technologies have proved to be in demand in many industries. The company found significantly greater demand from customers associated with warehouse logistics — retailers, postal and logistics companies that are actively working or at least studying the possibility of introducing unmanned loaders (Amazon is one of the pioneers of such solutions). An even larger group of potential customers are companies that need to know the exact location of employees. The Marvelmind solution allows you to determine the coordinates of an object in a room with an accuracy of up to centimeters, so you can configure the solution so that, for example, it signals if one of the employees is dangerously close to a working conveyor, etc. Railway companies, as well as companies whose employees spend a significant amount of time underground, such as mining, construction, etc., are showing interest in the Marvelmind solution.
“Determining the coordinates of people is still the most ambitious task with the maximum revenue potential for us. To work in this segment, we had to make a number of fundamental changes to the solution as a whole. Initially, the device emitting the signal was located on a moving object (for example, a robot), and the receiving sensors were on the walls of the room. In the case of people, for a number of reasons, it was more convenient to do the opposite: place only receiving sensors on a person, ” says Maxim Tretyakov.
In 2018, Russia accounted for only 13% of Marvelmind Robotics sales, and the largest market is the United States. The company supplies a software and hardware solution, its assembly facilities are located in China, and partners are engaged in the implementation — either the customers themselves or system integrators.
The company expects that the demand for its solutions will grow at a significant pace worldwide, including as labor automation increases. For example, according to the calculations of the American company Brooking Institution, in the transport industry, human labor can be replaced by machines and algorithms by 55%, in construction — by 50%, and in production-by 79%.