The future of law enforcement

We’ve taken steps in this column not to get dragged into the humans vs. machines debate. Many think the world is moving ever-closer to that so-called “singularity” in which machine intelligence exceeds that of humans.

We’ve always taken a different point of view: Let machines do what machines do well, and let humans do what humans do well. By working together instead of in opposition, we can create an amazing future.

Let’s look at one area in which that notion is already working its way into practice. It first appeared in 1987 as a campy cyberpunk movie, followed by a short-lived TV series. Its real-world incarnation might not look exactly like its fictionalized origins, but in some ways it’s remarkably close. Welcome to the world of … RoboCop.

Technology-augmented policing

Making RoboCop a reality means overcoming some longstanding challenges. But those deal more with intangibles than technology. As such, we’ll focus on the information processing and decision-making aspects of policing.

Whether out in the city streets or along the highways and byways of the rural countryside, officer discretion is a hallmark of law enforcement. First responders are called upon to serve, inform and make on-the-spot decisions that can end up doing either tremendous good or significant harm.

The means for guiding officer discretion has evolved over the years. Moving from mostly written directives to modern simulation-based training to the more recent use of body-worn cameras, sensors and computers, the volume and complexity of information keeps piling on. The Internet of Things (IoT) will only add to the problem.

But there are opportunities as well. Driverless vehicles will provide a significant advantage. Instead of worrying about traffic, first responders will use the time en route to review building plans and other relevant background information, all while keeping an eye on real-time video streaming from the destination site. The result: Officers will arrive on the scene better prepared.

Our vision for the future must support nearly 18,000 state and local law enforcement departments in the United States alone and extend across international borders. That means building intelligent systems that can rapidly ingest massive amounts of data from a variety of sources and apply the right decision protocols to guide officers and agents in the performance of their duty.

To illustrate, let’s look at a possible future scenario. Keep in mind that this is only one of countless situations a law enforcement officer might encounter on any given day.

A world in which anything can happen

A police officer in a driverless patrol car responds to a reported theft at a local chemical plant. Along the way, she reviews the protocol for investigating a possible chemical leakage or theft, which has been downloaded to her handheld mobile device.

Arriving at the plant, the officer opens her car door. That automatically activates her body camera, which begins streaming video to the Real Time Crime Center (RTCC), one of many data fusion centers located across the country. Once inside, the officer’s mobile device prompts her with a series of questions that follow different paths based on the answers of the person being interviewed. Her body-worn camera captures the whole process, all under the watchful eye of her supervisor.

One of the answers from the plant owner reveals that the amount of chemical stolen was enough to automatically trigger an alert to both the fusion centers and the FBI-led Joint Terrorism Task Force. That alert also places the hazmat team at the local fire department on notice for a possible response, just in case any leakages occurred or safety systems were compromised during the theft.

The responding officer receives information that there are cameras on site. AI-based software prompts her to obtain permission for the RTCC to access the business’ IP address. That helps to identify any suspicious people attempting to leave the premises undetected.

Meanwhile, analysts at the RTCC begin to gather information on the business, owner, workers and customers. They open a virtual collaborative workspace that is shared with other fusion and counterterrorism centers. Within minutes, one of the other centers reports that a similar theft occurred in a neighboring state a few months earlier.

Still being prompted through her mobile device, the officer asks the plant owner about any suspicious people who might have been lingering in the area and about employees who have not shown up for work. He says that one employee did not report to work, and he had noticed that in the past few weeks that person was on his smartphone a good deal. He gives the officer the employee’s picture and information.

The officer uploads her report. That automatically causes additional information about the subject and his network to be pulled from various data sources, including public records.

After closely monitoring the situation, the local FBI field office obtains a search warrant and heads to the missing employee’s last known address. The home is unoccupied, but agent-worn sensors help locate the stolen chemicals.

As it turns out, a few months prior to the reported theft, the National Counterterrorism Center (nctc.gov) issued an alert about a threat involving terrorist cells capable of preparing and using a certain type of chemical to make explosives. That particular chemical’s properties and unique markers were automatically downloaded to all body-worn sensors within a certain geographic area, including those worn by the responding FBI agents.

With the aid of live video streaming and facial recognition software containing the images uploaded from the police officer’s report, highway patrol officers apprehend the suspect near the state line. A terrorist attack has been thwarted.

In the early days of paper report filing and long decision processes moving up and down the chain of command, that scenario would have had little chance of a successful outcome. Only with tight integration of the full range of human expertise and supporting technologies can we hope to counter the complex dangers the future might bring.

A huge opportunity for KM

KM will contribute heavily to the evolution of law enforcement and related domains. A top priority is identifying specialized expertise and making it readily available. In law enforcement, everyone from citizens to headquarters staff needs to be ready to respond to a wide range of threats at a moment’s notice. Every possible scenario, whether a social network-incited riot or a hidden explosive device or a crazed SUV driver, brings with it the need for instant access to specialized expertise. That presents both challenges and opportunities.

Supporting activities from continuously monitoring, mining and interpreting massive streams of real-time data to determining the best course of action while anticipating possible countermoves is about as knowledge-intensive as it gets. From the army of investigators and analysts working behind the scenes to frontline responders, we need the speed and accuracy of machines combined with the very best in human expertise.

The threat is constantly changing and doesn’t care about obeying the law. That calls for continuous review, formulation and clear communication of policy and governance. That is especially true in law enforcement, which has come under intense scrutiny in recent years, as every move is subject to being second-guessed on the evening news.

There is also the ever-present need for standards. While the topic might sound boring, our future vision has little chance of succeeding if policies, processes and systems are not fully compatible and operating in sync.

Finally, opportunities for KM extend well beyond law enforcement. Why stop with RoboCop? Surely RoboFirefighter, RoboParamedic, RoboSoldier, even RoboQuarterback can’t be too far behind.