In the first part of this series of blog posts, I spoke about two aspects of the new land of convergence – RoBIoTICS and the convergence between robotics and biology domains. Today, I look at the convergence of robotics and the Internet of Things.
RoB-IoT-ICS – Convergence of Robotics and “Internet of Things” (IoT) domains
Internet of Things and Robotics are both highly synergistic domains. So what does the intersection of both mean?
Robotics and IoT are much more than connected devices. Robotics is actually a continuum: It moves from pure automation, which basically has devices sensing and acting, to robotics which connects and senses its environment and then acts in the physical world. Over the last ten years, a lot of communication technologies have been developed that directly supports the robotics industry sector. Some robotics sectors are actually using some direct research from communication domains such as facial recognition or natural language processing.
The impact of IoT on Robotics starts with the sensors. The more they proliferate in the environment, the more robotics will use these capabilities of sensing. Sensing with manipulation capabilities endows robots with specific capabilities (physical motion and movements), which set them apart from other automated systems. The IoT domain can support robotics to offer additional mobility and navigation features, and thus enable robots to move from one point to another.
Multiple new streams of research, currently underway, will soon allow devices the capability of knowing and self-describing themselves. Soon, everything from IoT will have a known location, and the environment will be mapped easily by robots. These maps will in turn be updated in real time by adding newer IoT devices. These advanced capabilities will make it easier for robots to move about in an unknown environment, locate robots wherever they may be, and assist robots moving from place to place.
Self-describing objects with tangible data structure will not only enable moving capabilities, but also object recognition and manipulation from robots. Robots need to know their environment, recognize objects in their environment in an autonomous way to be able to manipulate them. Visual Object recognition could be difficult to accomplish, as it typically relies on complex visioning systems, intensive 3D modeling processing, and generally web connectivity and cloud-based processing. Using descriptors embedded in real objects (or having objects with self-description capabilities able to communicate) could greatly improve object recognition by robots. This technique will allow robots to recognize objects easily, making it the first step to manipulate objects.
In addition, if an object has the capability to tell where it is and what it is then robots can use this information effectively.
Robots have always been highly “sensored”. They are able to fuse sensor information from a lot of IoT connected devices in their environment and pre-process data locally, before sending it to remote devices, services, and cloud-based service. But IoT can also use the robotics capabilities as an advanced type of sensing system: robots can be used as a sort of advanced sensors gateway with very powerful processing power for IoT devices, for example, by providing to IoT objects the connectivity up to the cloud and other distributed systems.
In addition many robots exist, and they are clearly social entities. They can sense and act in their environment thanks to IoT, but can also support face and speech recognition, or Natural Language Processing to act directly with humans, and in a near future, even be part of them.
It is clear that Robotics and IoT domains are fully interconnected, and enrich each other. We can easily envision what the future is going to be with biology, micro & nano technologies, and the future potential convergence in these domains. The convergence between humans and Robotics is clearly underway.
(To be continued)… watch this space for the next blog post!