As we look to the future of driverless cars and autonomous delivery drones, one of the most existential questions of their future is how to prevent them from being used for harm. What is to stop a driverless car from being loaded with explosives and sent off to its destination or a delivery drone from being loaded with explosives or a chemical weapon and used to target an individual with pinpoint precision? Terrorist groups are already making increasing use of modified civilian drones as weapons platforms, including as explosives delivery systems, while militaries are increasingly eyeing such drones as inexpensive and deniable weapons systems ideally suited for close-quarters navigation of urban environments. What can manufacturers do to curb such repurposing of their products?
A few hundred dollars, a few open source packages and an afternoon of work is all that’s required to turn an off-the-shelf civilian drone into a “killer robot” autonomous weapon today. Even less effort is required to jury rig a manually controlled drone to drop a payload and control it remotely by camera.
Driverless cars and trucks are still in their infancy, but as they become more commonplace, they post obvious risks given the long history of vehicles being repurposed as weapons. The ability to utilize a vehicle without requiring a human at the wheel would likely dramatically accelerate this practice.
What can companies do to curb such use of their products?
Driverless cargo trucks could utilize intensive cargo screening and automated alerts that would cause a vehicle to stop if it detected something being attached to it on the roadway.
Driverless cars pose a more complex challenge, especially as they eventually serve as delivery services. It would take little for a restaurant delivery to be swapped out for a more nefarious package. Weight and various kinds of sensors could help determine the composition of deliveries, but this would not stop a package from being attached to the exterior of a car while it waits at a red light or while the intended package is being placed inside. While there are potential solutions, including exterior cameras checking for unusual activity and interior sensors assessing the composition of a package, such systems will require novel development that is not yet part of the routine conversation about driverless vehicles.
Civilian drones pose the greatest immediate threat given that they are already being used in such applications by terrorist organizations and militaries are investing heavily in efforts to systematize their reuse as weapons platforms.
Drone manufacturers have responded by placing geofences around sensitive locations, but this only stops ordinary operation. An increasing array of projects, both open source and military, focus on replacing drones’ built-in navigation systems with external controllers that bring together a range of sensor platforms to offer autonomous flight and targeting, including facial and structure recognition.
There is little drone builders can do to curb such use, since at the end of the day a drone’s onboard controller can simply be swapped out and any integrated safety measures, like proximity kill switches could be bypassed.
Instead, the burden falls to society to examine ways to make our autonomous future safer. Land and air-based checkpoints could examine driverless cars and autonomous delivery drones to ensure both they and their cargoes do not carry any unknown or dangerous materials. Improved real-time tracking and wide-area scanning for dangerous goods such as expanded radiological detection and new forms of long-range chemical and biological weapons and explosives detection would allow autonomous vehicles to be remotely monitored, while improved aerial building security would help residences, businesses, prisons and other facilities secure themselves from rogue drone operators.
Drones have exposed the limitations of today’s ground-based security measures, while autonomous vehicles and drones threaten to shift the dynamics of conflict when offensive action no longer places human operators at risk.
In the end, as we race towards our autonomous future and pour all of our efforts into building self-driving and self-flying vehicles, we need to spend a bit more time examining the ways these new technologies could be used for harm and new approaches to mitigating that danger.
