Author: 

Main Authors: Sascha Griffiths

Adittional Authors: Stratos Arampatzis

Focus Area: 

Artificial Intelligence

Who stands to benefit and how: 

The paper is directly relevant for all parties interested in connecting clouds to systems which have physical components. Cloud robotics is not only about augmenting robots but thinking about larger ecosystems in which devices are connected and software and hardware components are closely intertwined. Therefore, cloud robotics is relevant to a large audience.

Position Paper: 

Cloud robotics is a means of both making robot software development more efficient and augmenting robot cognition. The idea of cloud robotics goes back to web-enabled robotics which was meant as an aid to autonomous robots in case their knowledge was insufficient to deal with a specific situation which programmers did not envision beforehand. However, for the past years advances in cloud computing have made the concept more powerful and cloud robotics now paves the way for robots to have advanced artificial intelligence at their disposal.

What is cloud robotics?

Cloud robotics comprises three elements: robotics, artificial intelligence and cloud computing. Before the paradigm was known as cloud robotics, it was basically called “web-enabled robotics”. The idea was that robots which encountered a problem which was beyond their programming would use a “search engine” to query the world wide web for them to obtain a solution. In this paradigm the robot would be mainly autonomous, and the cloud would return knowledge in a way that the robot can use and understand the information obtained from the cloud. Currently, most robots are not fully autonomous and generally artificially intelligent. Thus, cloud robotics is a means of using modern cloud computing to deploy artificial intelligence or in a broader sense software and knowledge onto robots. The cloud in this understanding becomes more than an occasional fix. Instead it becomes central to robot intelligence.

Modern artificial intelligence is used synonymously with machine learning and specifically deep learning. The large-scale use of deep learning was not made possible by significant advances on the algorithmic level but rather through progress in processing speed, larger memory, increased storage resources and the use of parallelization. Like one would not and essentially could not do deep learning directly on a mobile device or single workstation, this kind of resource intensive computing would not be done on a robot but rather delegated to the cloud. Thus, cloud robotics offers resources related to processing, memory and storage to a robot.

Often, cloud robotics is simply equated with “offloading heavy computation into the cloud”. This has several effects:

  • It makes robots lighter and reduces costs, by not having to have dedicated additional computer hardware close to or directly on the robot.
  • It makes robots faster and more efficient, by not having too many processes running in the background.
  • It makes robots more powerful or intelligent, by allowing robots to benefit from advances in artificial intelligence of the type described above.

At the simplest level, cloud robotics just makes a robot’s battery last longer through the previously named methods but at the more ambitious end robots will also become cognitively augmented. There is no need to develop software for a specific robot or a specific type of robot. Instead one can create artificial intelligence methods which are hardware-agnostic and are deployed on a larger number of heterogenous robots. This can, for example, be related to navigation, object or face recognition, or speech and language-related tasks. Apart from being connected to the cloud, such software can also be downloaded in the form of apps. In this way, cloud robotics is also a field concerned with new ways of deploying software onto robots. All of this will accelerate the development but also deployment of cognitive robots.

Cloud Robotics and Cyber-physical Systems in a Connected World

Like other physical and software components, cognitive robots will just become part of a larger landscape of cyber-physical systems. ‘Cyber-physical systems’ are all technologies in which software components and physical components are closely intertwined. Cloud robotics, in this sense, is a means of specifically tackling the connectivity of devices which are generally excepted as robots with larger ecosystem of artificial intelligence and the Internet interacting with humans and their environments. These environments are increasingly technology-centered while still being human-centered at their core. They include self-driving cars, drones, smart cities, smart factories, smart homes, the Internet of Things and a large variety of robots. The connectedness of all devices, some of which only live on servers and the internet and others in the physical world, will only increase. Cloud robotics is a means of connecting robots on an individual basis to the larger infrastructure of the connected world which new technologies are creating.

Year: 
2019