August 4, 2015 at 20:55IoR (Internet of robots) - a new concept, but the sores are the same
Hardly among the readers of Habr, and even more so GT it is possible to find a reader who has not heard about the Internet of things (IoT). At the same time, probably, most readers are already well aware of what it is, they understand how this concept is developing, what blessings will bring us in the future, and what are the difficulties and even threats.
However, I have not yet met the term Internet of Robots (IoR), or "Internet robots." As a representative of the Robotophob community, whose main mission is to convey to others that the development of robotics is not only universal happiness in the future, but also serious potential threats to the safety of people, which we should seriously think about now, in the early stages of robotics development, I will try to highlight the possible appearance of such a term, and also try to identify the advantages, and of course, the threats that this concept can carry.
Let's try to define IoR. By analogy with IoT, this is the concept of a computer network of physical objects (“robots”) equipped with built-in technologies for interacting with each other or with the external environment, eliminating the need for human participation from part of actions and operations.
Currently, we see that one of the main technologies for the development of robots will be cloud services. Cloud technology really helps a lot in the development of robotics:
- To be sufficiently independent of people, robots need to process a huge amount of information. In these processes, the clouds take on the main load, thereby cheapening and simplifying the hardware of robots (remember Siri, OK, Google).
- all updates of functionality, firmware, “training” and improvement of robots through cloud technology allow robots to improve without the need for their owners to participate in these processes, receiving new updates automatically (the so-called RobotAppStore).
- having the opportunity to learn independently, robots through cloud technologies will be able to share new skills and acquired knowledge, data that they can collect when interacting with the world around them, which will allow robots to improve and solve previously unfamiliar tasks without the participation of their creators and manufacturers.
- Well, in the end, if the robot gets into a difficult situation, he will be able to contact the “support hotline”, where he will be helped to fix the malfunction, download the necessary plug-in, etc.
Surely, not all the scenarios and opportunities that IoR gives us for the development of robotics are listed above, however, in my opinion, other aspects of this concept are no less important, namely, security issues. We regularly witness the emergence of new vulnerabilities, new threats in all areas of IT, whether it is hardware, software, services, IoT or even information protection systems. No matter how representatives of information security try, hackers are finding new ways to bypass and hack. Without thinking about security in a timely manner, we end up with serious losses, informational, financial, and reputational. Think, they are now hacking and gaining remote access to the systems of entire nuclear power plants, Ministries, data from top officials of states and spy services. What can I say about the devices of ordinary users.
And after all, many such incidents could be avoided by timely increasing the level of security, introducing protective measures during the design, and not imposing them after the fact.
We are talking about computers, servers, mobile devices, but let's imagine what can happen if there is an opportunity to get remote access to robots - industrial manipulators, cars, medical and military robots, assistant robots that will be among us and not just communicate, but they will also be able to act, affect the environment and people:
- robots equipped with a large number of sensors, video cameras, microphones, connected to the Internet in essence - an ideal listening and broadcasting device - a robot can end up in any room, premises, office necessary for an attacker ...
- having manipulators, a robot to which remote access with control is obtained is practically a weapon - it can easily cause direct and indirect physical harm to a person, switch some toggle switch off time, attack, open the front door in the end to the thief.
- connected to the Internet, the robot is, in fact, an open vulnerable system - an attacker can roll up incorrect updates, change operating principles, disable certain rules and restrictions, download a program or algorithms that are aimed not at leaving but at harm.
- united in a single network, robots can exchange any information that they can perceive, whether it is an image, sound, other data. Imagine how easy it is now to find any person to someone who takes possession of remote access to this network.
Again, you can list a huge number of negative scenarios that are already being implemented in IT or which can be implemented with the widespread dissemination of the IoR concept, if we do not start thinking about security in a timely manner.
Now there are practically no standards, let alone world standards, that would somehow regulate these issues. And it may happen that in the robots of the future even the rules of Azimov’s robotics will not be laid down - after all, development of robotics is important for everyone now, but we can think about security, as usual, later on?
However, I have not yet met the term Internet of Robots (IoR), or "Internet robots." As a representative of the Robotophob community, whose main mission is to convey to others that the development of robotics is not only universal happiness in the future, but also serious potential threats to the safety of people, which we should seriously think about now, in the early stages of robotics development, I will try to highlight the possible appearance of such a term, and also try to identify the advantages, and of course, the threats that this concept can carry.
Let's try to define IoR. By analogy with IoT, this is the concept of a computer network of physical objects (“robots”) equipped with built-in technologies for interacting with each other or with the external environment, eliminating the need for human participation from part of actions and operations.
Currently, we see that one of the main technologies for the development of robots will be cloud services. Cloud technology really helps a lot in the development of robotics:
- To be sufficiently independent of people, robots need to process a huge amount of information. In these processes, the clouds take on the main load, thereby cheapening and simplifying the hardware of robots (remember Siri, OK, Google).
- all updates of functionality, firmware, “training” and improvement of robots through cloud technology allow robots to improve without the need for their owners to participate in these processes, receiving new updates automatically (the so-called RobotAppStore).
- having the opportunity to learn independently, robots through cloud technologies will be able to share new skills and acquired knowledge, data that they can collect when interacting with the world around them, which will allow robots to improve and solve previously unfamiliar tasks without the participation of their creators and manufacturers.
- Well, in the end, if the robot gets into a difficult situation, he will be able to contact the “support hotline”, where he will be helped to fix the malfunction, download the necessary plug-in, etc.
Surely, not all the scenarios and opportunities that IoR gives us for the development of robotics are listed above, however, in my opinion, other aspects of this concept are no less important, namely, security issues. We regularly witness the emergence of new vulnerabilities, new threats in all areas of IT, whether it is hardware, software, services, IoT or even information protection systems. No matter how representatives of information security try, hackers are finding new ways to bypass and hack. Without thinking about security in a timely manner, we end up with serious losses, informational, financial, and reputational. Think, they are now hacking and gaining remote access to the systems of entire nuclear power plants, Ministries, data from top officials of states and spy services. What can I say about the devices of ordinary users.
And after all, many such incidents could be avoided by timely increasing the level of security, introducing protective measures during the design, and not imposing them after the fact.
We are talking about computers, servers, mobile devices, but let's imagine what can happen if there is an opportunity to get remote access to robots - industrial manipulators, cars, medical and military robots, assistant robots that will be among us and not just communicate, but they will also be able to act, affect the environment and people:
- robots equipped with a large number of sensors, video cameras, microphones, connected to the Internet in essence - an ideal listening and broadcasting device - a robot can end up in any room, premises, office necessary for an attacker ...
- having manipulators, a robot to which remote access with control is obtained is practically a weapon - it can easily cause direct and indirect physical harm to a person, switch some toggle switch off time, attack, open the front door in the end to the thief.
- connected to the Internet, the robot is, in fact, an open vulnerable system - an attacker can roll up incorrect updates, change operating principles, disable certain rules and restrictions, download a program or algorithms that are aimed not at leaving but at harm.
- united in a single network, robots can exchange any information that they can perceive, whether it is an image, sound, other data. Imagine how easy it is now to find any person to someone who takes possession of remote access to this network.
Again, you can list a huge number of negative scenarios that are already being implemented in IT or which can be implemented with the widespread dissemination of the IoR concept, if we do not start thinking about security in a timely manner.
Now there are practically no standards, let alone world standards, that would somehow regulate these issues. And it may happen that in the robots of the future even the rules of Azimov’s robotics will not be laid down - after all, development of robotics is important for everyone now, but we can think about security, as usual, later on?
