13-09-2021 | | By Robin Mitchell
The IoT revolution helped push electronics to new frontiers with the development of cellular IoT, self-powering devices, and mass data collection, but this came at the cost of large-scale security vulnerabilities. What does the next wave of robotic advancement offer, why should engineers take extra precautions with robotics and security, and what should engineers specifically worry about?
Modern robotics brings are awe-inspiring and complex machines with the ability to perform repetitive tasks in production lines, transport materials across sites, and operate 24 hours. Replacing human workers with robots provides a business with many benefits, including the lack of environmental controls, reduced wage prices, and increased productivity.
However, humans have only scratched the surface of what robotics can do, and the developments in AI could see a rapid acceleration of robotic integration into everyday life. From carrying luggage to helping out around the home, robotics could effectively eliminate the need for manual labour and remove the risk of injury to humans. This elimination of labour would further cause society to rethink what gives anything value and how resources should be distributed.
Of course, such robotics are far from being realised due to the complexities involved. Bipedal robots are still in their very early days, with Boston Dynamics has made a handful of robots that operate in carefully controlled environments. Four-legged robots may have proven to be operational, but even these still suffer from confusion and challenges that often require human input to correct. Even self-driving cars are still science fiction; current self-driving systems need a user at the wheel to take over during incidents.
While robotics is still in its infancy, engineers should start to think about the bigger picture and predict how their designs could shape future products. To understand why this is essential, one only has to look at the IoT industry to see what happens when engineers don’t plan ahead.
The IoT industry has grown rapidly, with more than 20 billion devices connected globally, and this growth has helped accelerate the development of SoCs, microcontrollers, and sensors. While this growth has helped improve internet-related technologies, the lack of foresight in security has created a somewhat unprotected ecosystem of devices.
This lack of foresight came about in the early days of IoT when engineers didn’t consider how many devices would be produced globally. Lack of consideration was also given to how those devices would have internet access, how those devices would have access to personal data, and the idea that their design was not worth protecting. Furthermore, many designers exercised poor security practices such as default passwords, unprotected network connections, and lack of encryption, making them incredibly easy to hack and abuse.
In robotics, engineers need to start thinking about how robotics will change the future, how they will be used, and how a future hacker may try and take advantage of the robotic system. For example, an attacker may use the robotics movement capabilities to injure a targeted individual or perform some action in a restricted area (such as unlock doors and/or windows). An attacker may also use robotic systems to spy via cameras and microphones in the same way IoT devices can be used to spy.
Firstly, robotic devices should be considered highly advanced IoT devices as they are almost always network controlled, provide multiple sensory inputs, and record data to some degree. As such, all the challenges and solutions faced by IoT devices should be applied to robotic systems. This includes using safe boot methods to prevent malware from being loaded, encryption to keep data safe, and no default passwords to prevent unauthorised entry.
However, robotics also presents the challenge of causing harm and performing actions that may be unwanted. Future robotic systems may run two independent computers; one for remote control and another for isolated control.
Simply put, the remote control computer is what controls the robot under regular operation (i.e. commands). Still, the second isolated computer runs on top of these commands to check for potentially malicious activities, including harming humans or accessing restricted areas. If any detections are made, it can override the primary computer and safely shut down the robot.
Overall, engineers will need to carefully consider how their robotic systems will be used in the future, what opportunities they will open up, and how to best protect against future threats.