Policing in the era of mobile robots

What happens when police encounter a ‘criminal’ mobile robot? How are police, for example, to stop autonomous vehicles? For many of us, the typical way to interact with the police is via traffic stops (Kerr, 2017; Woods, 2019). But, when we are passengers in a self-driving car that obeys traffic rules, there might be no need for such interventions. This scenario is particularly important for communities disproportionately targeted by discretionary policing practices (Joh, 2019). Fewer stops by police will certainly be beneficial for police officers as well, as it will lead to a de-escalation of risk associated with this type of police work. As Woods (2019) suggests, AVs will obey rules, making traffic stops less likely, decrease assaults against officers, and lead to an overhaul of drinkdriving and licensing laws and policies. Data transmitted to and from autonomous vehicles are likely to assist in investigations of traffic incidents when they occur. Some manufacturers have been developing systems that will communicate traffic violations directly to police, bypassing humans passengers (Joh, 2019). Developers have also been testing the tools to remotely immobilise self-driving cars from an emergency hotline. Just imagine the following scenario: your self-driving car is stolen. You quickly contact the police, and they request permission to take over the vehicle. They drive the vehicle to the police station, with thieves still inside (Joh, 2019). This ‘remote control’ policing could undoubtedly reduce the risk associated with high-speed car chases and arrests, both for officers involved and the passengers. Indeed, many police shootings happen during stop and search attempts or car chases by police (Joh, 2019).

Emergency services use teleoperated robots such as ‘public order drones’ (Hayes et al., 2014; Sandvik, 2016) or bomb defusing mobile robots for a range of purposes, such as surveillance, public order, safeguarding persons and objects, and preventing or detecting crime (Finn and Wright, 2012; Marin, 2016). A robot that was sent to a house to taser an armed offender under the direction of the supervisor prompted a debate on using robots to save lives (Cook and Zhang, 2020). Some researchers suggests that public acceptance of police drones is high (Boucher, 2016). Given that mobile robots and other technological advancements used by law enforcement integrate Al and the loT technology, future policing is likely to be big data analysis and intervention, rather than hands-on police approach (Kerr, 2017; Renda, 2018). Although ‘weaponised civil drones’ may or may not patrol our skies and roads in the future (for more context see Finn and Wright, 2012; Sandvik, 2016; Marin, 2016), policing will certainly change. It may well be offshored to businesses, agencies, universities, banks, and citizens. Rather than providing officers on the ground, customers will need to purchase the technology, affiliated (or not) with the local police department. One such example is an outdoor security robot K5. The machine that visually resembles Star Wars’ muchloved smart robot R2-D2 is deployed to patrol the streets, university campuses, parking lots, sports stadiums, hospitals, and other public spaces in the US (Markman, 2018). As the company claims on its website, the future is already here. Powered by AI, K5 promises to stop the crime before it happens, while wearing the official police logo, or acting as an independent ‘security guard’.

Like discussions vis-à-vis the loT and Al, it is dubious whether the advances of mobile robots will eradicate the danger of over-policing, discretionary policing practices, and pre-crime. The narrative of benevolent technology should be taken with caution: even if/when technology becomes fully autonomous, the question of foundational algorithms and limitations of machine learning remain. A vast amount of data collected and transmitted by mobile robots could be a vital investigative tool for law enforcement, as discussed in Chapter 4. However, they could, like scenarios investigated in previous chapters, lead to pre-crime. Moreover, growth and further developments in autonomous mobile robotics are a concern for crime control agencies. To date, drones have been used for stalking, intimidation of police officers and citizens, drug smuggling (especially to prisons), spying on homes and looking for potential targets for burglaries (Kogers, 2019). In the future, this list is likely to get longer. In December 2019, police struggled to disable drones that interrupted air traffic at Gatwick airport in the UK. For three days, over 29 drone sightings caused a significant disruption in air travel just before Christmas (Rogers, 2019). With more drones used for commercial purposes and deliveries, police intervention might be necessary more often’. The question we need to ask is how will the police engage with disorderly and criminal behaviour of mobile robots and what will the impact of such interventions be, especially on vulnerable and marginalised social groups (Finn and Wright, 2012). We need to be wary of the ‘mission creep’ that might see the increase in police use of autonomous mobile robots for other purposes.

Swarms of drones powered by Al that communicate with one another are already a reality. Soon, drones that work as a team will make informed, and to some extent, autonomous decisions based on data they collect or receive from each other, the environment, and human instructors. They will have autonomy in reaching set goals and might create new goals. We need to be prepared for the scenario that some drones or AVs in a swarm, for example, might distract the attention of police while others from the pack offend. We have seen an example of this in the 2019 attack on Saudi oil fields, where several drones acted as a decoy while others committed an attack that was a significant blow to the Saudi economy, but also their US protectors (Rogers, 2019). Will police resort to hacking or other denial of service interventions to counter crimes committed by or via autonomous mobile robots? Critically, what new police powers will unfold, promoted as necessary to combat a new threat?

< Prev   CONTENTS   Source   Next >