Cloud robotics – coming of age?
Cloud robotics has been one of the most researched topics in the robotics space for the last few years. There is enough reason for enthusiasm. The possibilities of new business models and faster integration of AI and remote management capabilities have led to new ways of deployment of robots with customizable software. The sector has attracted all the major cloud providers like Amazon, Google and Microsoft who envision robots as the next sector to use the cloud backbone.
Amazon has its RoboMaker platform, which mainly focuses on ROS based development with simulation so that the development can be done iteratively. But, the solutions now also consists of over-the-air updates of software and easy deployment of robots on-site. Google is developing its Google Cloud Robotics Platform that combines artificial intelligence (AI), cloud computing and robotics whose aim is to enable “an open ecosystem of automation solutions that use cloud-connected collaborative robots.”
Similarly, Microsoft also is planning to get its Azure and Windows platform to robotics. Robotics-as-a-service (RaaS) is the new business model, robotics companies want to deploy the robots. This lowers the upfront cost for the customers and also gives leverage for the companies to generate service revenues with upgraded solutions and a large amount of data of operation helping in creating better robots. Other companies such as Honda, IBM, Huawei are also coming up with their open platforms to have a foot in the robotics market.
Some market estimates show a robust 20-30% CAGR for cloud robotics market for the next 5 years. The incumbent industrial robotics players like KUKA, Yaskawa, ABB have also emphasized their research in the area of cloud robotics and it is getting integrated into many of their products. But, using OTA updates and tracking maintenance data is only scratching the surface of the possibilities presented by cloud robotics.
Artificial intelligence at scale can be incorporated in a fleet of robots using a cloud backend. This is probably the most interesting application that could be possible if the cloud – robot communication is set up reliably. The robots could act as sensory data providers and much of the heavy-duty, AI-based data analytics can be done on the cloud. This allows the hardware setup at the robot to be cheap and minimal. Cobalt Robotics is one such startup which patrols building with robots for guarding premises and sends the streams of data to the cloud where AI-based algorithms can derive insights into making security operations better. Similarly, a few warehouse robotics startups are trying to sell a robotic fleet as complete automation.
Taking this one step further, research is focusing on using the cloud infrastructure for transfer learning in a fleet of robots. This means that while training robots in a particular task using AI, the learning from one robot could be transferred to a cloud pool and other robots can then learn from the pool as well. This could be a new paradigm in much faster training for artificial intelligence and many complex applications can be developed for robots. Rapyuta Robotics, a startup spun-out from ETH Zurich and based in Japan is working in this domain. The big validation came for Rapyuta as Microsoft Japan, announced a partnership with the company for developing its cloud robotics platform on the Azure platform.
Cloudminds, the pioneer in the cloud AI architecture for robots and business announced recently that it has received the order to supply over 10000 intelligent robots to Jin Yu Ao Environmental Technology and Zhongtai Min’an Security Services Group. Jin Yu Ao is a large scale cleaning services provider which is also heavily investing in the next generation 5G smart cleaning command center along with the Cloudminds technology.
Zhongtain Min’an, on the other hand, is a security services and property management consultant. Cloudminds technology would be used here to make smart decisions in the cloud for cleaning and advanced property management solutions. Cloudminds, which is backed by Softbank, is also rumoured to be planning a $500 million IPO. Much of this shows the confidence in the growth in this sector. Acquisitions have also started in the space. Recently, Formant acquired Formation to integrate teleoperation into robotic fleet management.
We believe, this kind of news will get prominent in this as well as the next year. With the 5G adoption becoming more widespread, the true potential of the cloud robotics technology will be realized. We would see robotics-as-a-service dominate the robotics market in the coming few years.
To deep dive and stay continuously updated about the most recent global innovations in Robotics and learn more about applications in your industry, test drive WhatNext now!