14Criminology· UX Team

Robocop 2.0

DisruptionProbability 70/100

// The science behind it

VISION-

The future advances in fighting crime can be divided into three categories: The judiciary and criminal law administration, Policing, crime detection and prevention and Crime investigation, criminology, and forensic science.

The judiciary and criminal law administration

Thought the years, government bodies in charge of the law enforcement had to find ways in which they’d cut the costs and let some room for technological advances to enter the crime fighting arena. To cut those costs, most governments will implement “smart courts” by 2050. Smart court implies AI-powered judges listening to cases over digital court hearings. The judges themselves appear as holograms which are synthesized 3D images of different judges. Defendant, witnesses, victims, lawyers, and the jury, although real humans, have an option to attend the court as holograms as well.

The hologram and AR technology will ultimately help remove the disadvantages of videoconferencing. A video tends to make people perceived as abstract, less of a participant and less relatable.

Courtrooms can feel very intimidating. When entering the waiting area, the only available seat is the one next to your assaulter. And the distance doesn’t get much greater upon entering the courtroom either. These technologies will therefore make victims more likely to come forward, as they no longer must face their assaulters in person if the case gets taken to court.

Those who do decide to attend the courtroom in-flesh will be greeted by a smart court robot guiding them to their appropriate seating locations.

There are no more stenographers. Court hearings are being fully recorded and searchable through via AI generated scripts. Everyone having access to the court documents can view the hearings as they happened, the courtroom layout and the participants in AR at the comfort of their homes.

Policing, crime detection and prevention

Parabolic microphones helping to detect someone far beyond the normal range of human hearing will still be widely used, but the way sound is used to policing advantage will expand even more. Specifically, gunshot detection via sound detecting sensors will become the main tool for police department collaboration, knowledge exchange and data collection for deep learning.

The sound sensors are being placed on corners of the buildings throughout neighbourhoods. If there’s shots fired - they detect, record, and timestamp the sound. The position of the gunshot is determined based on how long it takes for the sound to reach the sensors. Besides the sound location, the AI can also determine which firearm has been used based on the sound database. The data is then sent to nearest police department, and the scene inspection task is delegated to the officers in charge of patrolling that neighbourhood.

Compared to traditional hotspot analysis, which includes police vehicles patrolling the previously high-crime areas, this new approach minimizes over-policing, results in faster response, reduces risk for the police officers and increases the neighbourhood safety.

Deep learning of the gunshot data, crime locations, times at which they happen as well as the most common weather conditions when crimes occur will enable effective city mapping. City mapping will hence be used to predict future crimes. AI neighbourhood police cameras will detect the unpredicted crimes as they’ll be able to distinguish criminal activities from basic ones. Data on the type of crime committed will be automatically shared with the patrolling vehicle nearest to where the camera detected the activity. AI police cameras will also be common at the borders and airports. They’ll be scanning the premises for a match in their databases with photography’s of the Interpol wanted criminals. Another content database will contain child abuse images, which AI will use to search through users content and look for a possible match. The AI won’t be able to see the users’ content unless its in the prohibited database.

High-speed chases were of concern, particularly in the US, resulting in more bystanders and passengers being killed by police vehicle pursuits every year. This is why GPS darts, first introduced in the 2010’s, made a great comeback in the 2030’s onwards, and the problem of them not sticking on wet and cold surfaces was resolved. The darts are ejected from the police vehicle and glued on the surface of the pursued vehicle. The police vehicle could then slow down and notify the first police vehicle in the opposite direction of the runaway vehicle’s whereabouts.

Police vehicles got a revamp too, with driverless autonomous police cars becoming more prevalent. Speeding cameras and sensors weren’t dispersed adequately enough to catch all the people going above speed limit. Therefore autonomous police cars with built in speed scanning systems will patrol the streets, measuring speed of the vehicles and automatically creating the tickets to be sent out to offenders’ addresses.

The police departments will get another powerful backup from the robot police officers and dogs. Especially in high-risk situations where bombs and hostages are involved. The robots will have the Range R radar installed, allowing them to detect motion through solid walls. The robot police dogs, because of how compact they are, will be the main tool used in bomb deactivation. In case the bomb does deactivate – the only ”life” endangered would be that of a digital pet.

The accessibility of police will be improved via police reporting 3D chat booths. One policeman will remotely take reports from booths he’s in charge of. For example, a policeman might be working in one spot of the city in-flesh – but be present in multiple neighbourhoods where citizens can enter the booths and report the crime live.

In 2020’s, it was common for parents to be able to track their child’s phone location or for users to set up emergency calls. However, what to you do when you can’t track your children anymore once they’re of legal age? Or what do you do when your phone has been taken by the abductor or you can’t use it to activate the emergency call due to possible incrimination?

Crime investigation, criminology, and forensic science

Traditional 3D crime-scene imaging method is now elevated with the holograms, allowing holographic crime scene and event reconstructions. After the scan of the crime scene has been done, projectors are being placed around it, showcasing how the events played out with the holograms of the perpetrators and victims involved.

A once-known-as lie detecting pseudoscience using polygraphs is now being admitted in the court as evidence. In the previous decades, rise of the cheap computing power, brain-scanning technologies and AI has given birth to a powerful new generation of lie-detection tools and interrogation techniques. Rather than using Comparison Question Test (CQT), they’re now using Concealed Information Test (CIT) that doesn’t look for lies but memory detection. Memory detection is detecting whether the respondent has guilty knowledge of the criminal event based on his reactions, heart rate etc. An example of such test would be presenting the respondent with series of weapons and seeing whether they react or how they react once they’re presented with the weapon found on the crime scene. Traditional polygraph equipment using the finger sensors will be replaced with using motion sensing suits catching and analysing body movements. Together with AI, brainwave monitoring and motion suits, polygraphs will be used to quickly detect actions translating to crime involvement.

Catching criminals via DNA will also get significantly easier. Personal genomics and biotechnology companies will demand from people wanting heritage read out from their DNA to accept their DNA to be uploaded to Police department databases automatically.

PROOF

China has been using AI judges and holograms in courtroom since 2017, so the idea of smart courts isn’t very new or unlikely. In December 2019, China has announced that millions of legal cases are now being decided by “Internet courts” that do not require citizens to appear in court. A Hangzhou court official told China’s state-run CGTN television network that the smart court system operates 24 hours a day, seven days a week. Some of China’s courtrooms have also already incorporated robots welcoming and guiding participants.

China is not the only one to acknowledge the AI judges. In 2019, Estonia announced AI judge incorporation in small claims court to clear court backlog. The idea reportedly came after the Estonian Ministry of Justice requested that its chief data officer Velsberg Ott's assist in designing a “robot judge” that could adjudicate small claims disputes of less than 7000 euros.

We also have some inclination on the disadvantages of videoconferencing over holograms. The mediating feel of a video link and the absence of the complainant or judge in the courtroom creates distance between them and the jury. The jury is less likely to hence have sympathy for the witness. Nonverbal communication plays a big role in various criminal cases, like those of assault nature. Jurors use intuition to base their conclusions on, and this intuition might be compromised because of a choppy stream.

Stenography is a well-paid job, especially when its government related like in court. However, it won’t take long until governments realize how easily the costs can already be cut by replacing them with AI. AI script makers have become so good that we can already search through moments of streams or YouTube videos via the real-time generated script.

Gunshot detection is something that’s in use, but not by many police departments. ShotSpotter is gunshot detection technology that uses sophisticated acoustic sensors to detect,

locate and alert law enforcement agencies and security personnel about illegal gunfire incidents

in real-time, less than a minute after the shots have been fired. The only deficiency of the system is the fact that the sound first gets sent to ShotSpotter’s review centre where acoustic experts determine whether it’s a gunshot or, for example, a firecracker or something else. This is expected to be automatically recognised by the AI in the future.

AI could also make its way into neighbourhood cameras. In 2021, Facebook announced Ego4D, a project aimed at teaching AI systems to comprehend and interact with the world like humans, from a first-person perspective. This means that the AI will be fed with data sufficient to make them able to group activities into sports, shopping, gardening baking and other. Question is how long it will take the AI to be able to distinguish criminal from harmless activities.

UK Police is already taking advantage of the predictive crime mapping using the PredPol system. Faced with the report that found that the British police have wealth of data but lack of capability to use it, they decided to put it to good use and pinpoint high-crime areas, time and days of the year when the crimes happen in order to monitor them and deploy additional resources accordingly.

Child abuse detection on users of Apple products has been announced and delayed for two reasons. First one being that users do not understand that Apple is not able to view any of the content on users’ phones unless the illegal one matching their child abuse imagery database. The second opposing the tool are the abusers.

While the version of facial recognitions of criminals mentioned in the article is far less privacy intruding, since there are cameras on the borders and airports anyway, there are softwares that are far more dangerous and in use by police forces around the world. Clearview AI is a software that collects billions of images from Instagram, Facebook, Twitter and YouTube, among others, without being asked, and can then links images from sources like a surveillance camera to an identity or profile on social media. Most e of the Police forces using the tool are stationed in the US. Notably, Belgian police has also been accused of using the tool illegally in 2021.

According to USA Today, nearly every day, someone is killed during a high-speed chase between police and a suspect. Almost half of the 11,500 people killed since 1979 have been bystanders or passengers in fleeing cars. GPS darts exist and present somewhat of a solution but are yet to be widely enforced. Those opposing the darts argue that they’re useful only under special weather conditions where the glue can stick to the frame of the car.

Autonomous vehicles are a hot topic, but there haven’t been many talks of an automated police car. But in 2018, Ford submitted a patent for the autonomous police cars. Ford made a statement that even if the patent is approved, it does not ensure that a product will be produced. Based on the questionable future of autonomous police vehicles, it’s hardly possible we’ll see them patrolling the streets and scanning the license plates of vehicles driving above limit.

However, something on the topic of scanning is being implemented in Brussels, Belgium. Brussels scan cars driven by officials scan cars to check validity of the parking rights of parked cars. Instead of tickets on the windscreen, letters will be sent directly to the address of the offender.

Though we still don’t have robotic policemen, Boston Dynamics dogs are already finding their way into police departments. They were used by the Honolulu Police to scan homeless people for fevers and integrated by the Massachusetts State Police into the unit's bomb squad. In October of 2020, the NYPD police department used the dog to find a gunman who’d barricaded himself in a building after he’d accidentally shot someone in the head during a parking dispute in Brooklyn.

Other robots like those from Endeavor Robotics, Knightscope, SMP Robotics and Northrop Grumman Remotec are meant to help police departments via monitoring and surveillance.

Some cities’ police departments are very welcome to the incorporation of robots. Dubai is adamant on becoming the world's safest big city. In October 2021 Dubai Police signed an agreement with UiPath – a leading global software company - to enhance cooperation and collaboration in robotic process automation (RPA). In 2017, Dubai expressed hopes that robots will constitute 25 percent of its police force by 2030, with the next stage being to use them as receptionists in police stations.

In February 2020, Belgium unveiled the first virtual police counter, similar to the police 3D chat booth mentioned in the article. A 3D-projection of a police officer is filling in for a flesh-and-bone policeman.

3D scene imaging is something that revolutionised the crime scene investigation. Holograms are yet to receive their crime case application but have been imagined in movies like 2018 Kin or 2013 Iron Man. We could imagine crime scenes reconstructed in VR as well, but we chose holograms for the convenience of being able to collaborate with multiple experts at the same time without the need of bulky VR devices and the cost effectiveness of not having to completely reconstruct the entire crime scene virtually. Instead, holograms place the crime perpetrators and victims in the existing, real space, with blood splatters and room arrangement as they really are.

The polygraph dilemma on whether they should be admitted in court last to date. However, there’s no question on whether people suck at identifying the lies – because they really do, far more than the polygraphs. But there has been recent research in favour of polygraph accuracy when conducted in addition of the CIT or Concealed Information Test. Japan is so far the only country where the polygraph with the CIT is widely applied to criminal investigations. CIT has not yet been combined with motion suits, but the suits are being tested. Some scientists have put volunteers in the Xsens motion sensing suits and could detect liars at a better rate than with the traditional finger sensors based on how much they froze, fidgeted, or moved their joints.

Personal genomics and biotechnology companies have already helped solve a significant number of cold cases. In 2019, a woman shared her DNA with 23AndMe to find out more about her heritage. After some research she found out that she can share her results with GEDmatch, a free open-source DNA database accessible to law enforcement without the court order. Her DNA helped charge and arrest a man with the 1980 first-degree murder and second-degree kidnapping of Pruszynski.

The “Golden State Killer” was also caught with the help of GEDmatch and geniology companies. But the way they did it is that police used the DNA data and uploaded it themselves to undisclosed genealogy website (presumably one of the top ones like 23AndMe, AncestryDNA, MyHeritageDNA). This sparked some privacy concerns amongst genealogy website users, but the general consensus was that people were happy to know that their DNA data lead to catching a serial killer.

FEASIBILITY SCORE 70/100??

Though already existing, it’s hard to predict how feasible wide implementation of these technologies is because we don’t know at which pace the technology will become more affordable and accessible.

Most tech companies produce slowly, with high production costs, resulting in costly output. Instead of innovating the production processes, they move towards innovating the product itself or creating new products. Take for example the Boston Dynamics robot dog. At a hefty price of 75 thousand dollars per pup, it’s hard to imagine police departments willing to risk it blowing up during a bomb deactivation process.

Though the dog is getting more impressive by the year – the hefty price is leaving its potentials undiscovered. Chinese start-up Weilan has identified the cost problem and has, instead of trying to build a better robot dog, improved the ways in which its being produced. Their AlphaDog comes at a significantly lower price 2,5 thousand dollars.

The tech we’ll definitely see implemented more are the AI powered judges for small dispute resolving, AI stenographers for quicker record making and holograms and 3D projections for crime scene investigation. We’ll also see the rise of advanced law enforcement data collecting platforms used to solve crimes with deep learning and AI, because the data necessary for these advance platforms is already being collected – but separately and by different entities who need to find the way to collaborate together in order to better the society.

Whether or not we’ll see even more camera surveillance on the street will depend on how comfortable people will be with it, knowing possible disadvantages. People are more than ever worried about their privacy, and we see this trend continuing into other decades.

In general, the society fears that those against them setting boundaries are the ones profiting from them having none. But if by profiting we envision police departments catching criminals more efficiently, people might be more willing to support the cause, just the way they already do when they indirectly solve the crimes with their own DNA.

RESEARCH

China using AI judges and holograms in courtroom since 2017

https://www.lexisnexis.ca/en-ca/ihc/2020-02/robot-justice-chinas-use-of-internet-courts.page

Child abuse: Apple urged to roll out image-scanning tool swiftly

https://www.theguardian.com/global-development/2021/sep/17/child-abuse-apple-urged-to-roll-out-image-scanning-tool-swiftly

European Parliament authorizes online message filtering to detect child abuse

https://www.euractiv.fr/section/economie/news/new-eu-law-allows-screening-of-online-messages-to-detect-child-abuse/

Belgian police illegally used facial recognition software

https://www.brusselstimes.com/news/belgium-all-news/188743/belgian-police-illegally-used-facial-recognition-software/

Robot gives guidance in Beijing court

https://www.chinadaily.com.cn/china/2017-10/13/content_33188642.htm

Ray Worthy Campbell (2020), Artificial Intelligence in the courtroom: The delivery of justice in the age of machine learning

https://ctlj.colorado.edu/wp-content/uploads/2020/08/2-Campbell_06.25.20.pdf

Carolyn McKay (2018), The Pixelated Prisoner

https://www.taylorfrancis.com/books/mono/10.4324/9781315111506/pixelated-prisoner-carolyn-mckay

Vicky De Mesmaecker (2014), Perceptions of Criminal Justice

https://www.routledge.com/Perceptions-of-Criminal-Justice/Mesmaecker/p/book/9781138961289

Boston Dynamics Robotic Dogs Used By Honolulu Police To Scan Homeless People For Fevers

https://www.wbur.org/news/2021/07/30/boston-dynamics-robots-honolulu-police

The NYPD is sending its controversial robot dog back to the pound

https://www.theverge.com/2021/4/29/22409559/nypd-robot-dog-digidog-boston-dynamics-contract-terminated

Are Police Robots the Future of Law Enforcement?

https://builtin.com/robotics/police-robot-law-enforcement

The Range-R sends out radio waves that can detect the slightest movements, including breathing, from as much as 50ft away

https://www.bbc.com/news/technology-30904218

The Boston Dynamics Robot Dog Has Joined a Bomb Squad

https://www.popularmechanics.com/technology/robots/a29994082/boston-dynamics-spot-bomb-squad/

How to set up emergency location sharing on Android and iOS

https://www.theverge.com/2019/3/18/18267500/how-to-set-up-emergency-location-sharing-android-ios

GPS darts help stop high-speed police chases

https://www.news5cleveland.com/news/local-news/oh-cuyahoga/gps-darts-help-stop-high-speed-police-chases-local-departments

Ford wants to patent a driverless police car that ambushes lawbreakers using artificial intelligence

https://www.washingtonpost.com/news/innovations/wp/2018/01/30/ford-submitted-a-patent-for-an-autonomous-police-car-the-u-s-government-just-approved-it/

ShotSpotter Precision Policing Platform

https://www.shotspotter.com/

How technology is allowing police to predict where and when crime will happen

https://www.independent.co.uk/news/uk/home-news/police-big-data-technology-predict-crime-hotspot-mapping-rusi-report-research-minority-report-offenders-risk-a7963706.html

Holographic Crime Scene Reconstruction in movies

https://www.youtube.com/watch?v=zwLOflhZOBg

https://www.youtube.com/watch?v=PWgza1YQOfI