Technology is moving at a very fast pace and has reached a stage where artificial intelligence and analytics will drive the future. Humans are irreplaceable when it comes to analytical and intellectual capability but an amalgam of emerging technologies has brought machines at the same level and a self-driving car is a perfect example in this context.
A self-driving car is capable of sensing its environment and navigating without human input. These cars use a variety of techniques to detect their surroundings, such as radar, laser light, GPS etc. Advanced control systems interpret sensory information to navigate their path to the desired destination. These cars generally use localization and mapping algorithms also known as deep learning algorithms which receive data from multiple sensors to detect and track surrounding objects. Below are the five levels of self-driving cars:
- Level 0 (Zero Automation) – Driving as Usual. A human driver is required to operate the vehicle safely at all times.
- Level 1 (Driver Assisted/Function-Specific) – Intelligent features add a layer of safety and comfort. A human driver is required for all critical functions.
- Level 2 (Partial Automation/Combined Autonomous Functions) – Key automated capabilities become standard but the driver is still in control.
- Level 3 (Conditional Automation/Limited Self-Driving) – The car becomes a co-pilot. The vehicle manages most safety-critical driving functions in known environmental conditions. A human driver is still present.
- Level 4 (High Automation) – Capable of performing all safety-critical driving functions while monitoring environmental conditions. This is considered as full self-driving automation. A functional driver cockpit is still in place.
- Level 5 (Fully Autonomous) – Vehicle is completely driverless. These vehicles do not feature driving equipment.
It is estimated that around ten million self-driving cars will be on the road in next five years and at Level 5 (Fully autonomous). This technology has a number of potential benefits both socially and economically like reduction in traffic collisions or increase in road safety leading to lower insurance costs. Big names from the industry are taking a shot at developing this future technology which will change the dynamics of the automotive sector. Stakes are high as with any revolutionary technology and it is too early to say whether this technology will be a blessing or a curse.
But it is important to discuss the dark side of such a technology as there are a number of concerns in the adoption of self-driving cars e.g. liability concerns, safety concerns, legal and regulatory concerns etc. In addition, these cars are not sophisticated enough to parse through inconsistent road signs, volatile weather and unpredictable pedestrians and drivers.
Among all these concerns and challenges, there exists one dimension of risk which is different from the rest and needs to be addressed completely prior to the roll out of this technology commercially. I hope you have guessed it as we are talking about the cyber security risk surrounding the driverless cars i.e. whether these cars can be hacked and misused by bad actors.
According to industry experts if the sensors in the car being used for complete navigation are connected to the internet directly or have an interface to the internet it might be susceptible to hackers. A self-driving car is nothing but a collection of different types of complex software and anybody with some knowledge of IT knows that no software is completely free of vulnerabilities.
Therefore, the software governing these self-driving cars might contain vulnerabilities that can be potentially exploited by bad actors. Few examples can be, tricking a machine learning system like this self-driving car by changing a road sign to something else or hacking the car and taking its control to misuse it.
Another example can be related to data privacy where bad actors hack into the car to access your location, driving patterns and your favorite hot spots and misuse this data for their personal benefit or infect the car with ransomware to shut it down till you give in to their demand. Government agencies need to lay down strict guidelines for the adoption of self-driving cars including secure coding practices to minimize cyber security risk.
Countries like US and UK already have such guidelines in place for self-driving cars which include issues like cyber security. As it is said that a building is strong if its foundation is strong, same analogy applies in the case of complex technology like this one. Therefore, it is important to go back to basics of secure system development lifecycle for developing the car software.
As technologies tend to become more and more complex in the coming future, we will only be able to address all security related concerns by holding the ground with a strong foundation of security principles.
