ALPR: a new philosophy of best practices

Automatic License Plate Recognition (ALPR) has undergone a major evolution in recent years. Among the most significant changes the technology has seen has been a migration from closed hardware-dependent solutions to software-only systems with open platforms that allow for easy integration with existing surveillance cameras and other third-party security products. These new solutions also boast extremely advanced new capabilities in Artificial Intelligence (AI) and machine learning for more comprehensive and accurate data gathering and analysis. This new approach necessitates a new philosophy of best practices for ALPR usage. Following are some of the most crucial points in this new philosophy that must be considered as ALPR, and surveillance technology in general, move into a new future.

Fixed vs. Mobile

For most of its existence, ALPR has been primarily a mobile technology, consisting of cameras mounted on the exterior of patrol vehicles connected to a processing box and in turn to a laptop computer. Despite advances that have moved ALPR away from this configuration, there is still a tendency among many in the security industry to perceive it this way. While mobile ALPR is a useful tool up to a point, it has several significant drawbacks.

First and foremost, mobile ALPR has a limited coverage area. Put simply, the technology can only capture vehicles that happen to be near the patrol car. The cost of most mobile ALPR solutions is such that the vast majority of law enforcement agencies can only afford one or two systems at most. These two factors greatly reduce the chances that a given person of interest will be spotted by ALPR.

Second, mobile ALPR can place an inordinate burden on patrol officers, who are already busy and required to be focused at all times, by requiring them to learn and operate new technology. Any benefits that officers may gain from it may therefore be outweighed by undesirable distractions that could potentially put their safety at risk.

During the past several years, cities across the United States have been adopting citywide surveillance systems consisting of networks of fixed-location cameras deployed at strategic intersections and other points of interest. These systems often serve as the cornerstone of centralized Real Time Crime Centers (RTCC) that constantly analyze incoming data to provide actionable intelligence to law enforcement. ALPR would seem a natural fit for such a system, but it has often not been practical until now due to the cost of the hardware involved and the time and complexity required for its deployment.

The rise of flexible, software-only ALPR solutions that work with existing security cameras, are easy to deploy, and are cost-effective has changed that paradigm. With such technology available, cities can quickly and affordably integrate ALPR capabilities into their citywide surveillance networks without having to “rip and replace” hardware. This enables ALPR to be deployed on any camera or group of cameras the authorities may designate, providing constant 24/7 coverage, not limited by the location of patrol vehicles. Additionally, cameras in the system can perform multiple functions, continuing to monitor and record at the same time they perform ALPR.

With a simple viewer application installed on their car laptops, patrol officers can now receive real-time feeds of vehicle capture data from anywhere in their jurisdiction without having to learn a new system. In addition to the safety benefits officers can realize from such a solution, it also becomes far easier for authorities to coordinate searches for persons of interest from a central location.

Best practices, therefore, favor fixed-location ALPR solutions over mobile.

Data Storage and Sharing

There has been a great deal of controversy in recent years over ALPR’s ability to store large amounts of vehicle capture data. The concern, as voiced by organizations such as the ACLU and others, is that this capacity gives law enforcement an inordinate ability to track the movements of innocent motorists over a period of weeks, months, or even years, thereby violating their privacy. This argument is reinforced when one considers that some organizations not officially affiliated with law enforcement have established central ALPR databases capable of holding many billions of captures.

While it can be argued that law enforcement has neither the time nor the personnel to track motorists unless they are already criminal suspects, it is nonetheless important at least to consider the potential for misuse that such centralized data storage presents. This is especially true when the data is in the hands of non-law-enforcement organizations. The remedy usually proposed by civil rights advocates involves legislation at the local, state, and federal levels regulating how long vehicle capture data may be stored and under what conditions it may be kept longer than the regulated time and/or shared among various law enforcement agencies.

It is also important, however, for members of the security industry to make recommendations of best practices to assist legislators, vendors, and users alike to maximize the effectiveness of ALPR without violating personal privacy.

Reasonable regulations of the technology are certainly necessary and beneficial, not only to ensure proper use, but also to help instill public peace of mind. Some municipalities are leading the way in this effort; for example, New York City officials recently introduced for consideration a bill known as the Public Oversight of Surveillance Technology (POST) act. It is also important, however, for members of the security industry to make recommendations of best practices to assist legislators, vendors, and users alike to maximize the effectiveness of ALPR without violating personal privacy.

Most important in this regard is the necessity of keeping the data squarely in the hands of law enforcement. Many states have established “fusion centers” that are designed to aggregate criminal data statewide to assist in law enforcement investigations, and it would seem obvious that ALPR would be a natural dataset in these aggregations. It is interesting to note, however, that of the states that have such centers, a significant number have not added ALPR information to their databases. It is vital, therefore, not only to encourage more states to implement fusion centers, but also for the ones that exist to collect and store ALPR data. As more cities adopt mass-deployed ALPR as part of their overall security infrastructure, this would be a natural step. However, governments can also encourage this model of data storage through appropriate legislation.

At the same time, new advancements in ALPR are allowing capture data to be shared with law enforcement agencies locally. Through secure peer-to-peer networking, data can be sent directly from a city based entity to law enforcement to alert them of possible wanted persons on the property. Such technology also enables law enforcement agencies to share the data with each other in order to locate the movements of persons of interest. This data sharing happens in real time as well as being secure. Local sharing of data also has the advantage of helping to ensure that such data is as up to date and reliable as possible. When data is drawn from a large centralized database, it is often difficult for law enforcement to know how current and accurate it is.

Moving Forward

The security industry has changed, particularly regarding ALPR. New capabilities in AI and machine learning, built into more cost-effective and adaptable solutions, have rendered the technology far more than just a “niche” application. As more citywide surveillance systems that include ALPR are adopted across the country and around the globe to the necessity of a new philosophy of best practices for this vital technology is clear. This new philosophy is aimed at expanding ALPR coverage area, enhancing law enforcement officer safety and efficiency, ensuring the currency and integrity of data, and protecting personal privacy.

PlateSmart is setting an example through its development of software-only ALPR solutions with advanced AI and machine learning. The additional data-gathering capabilities of these solutions have spearheaded the evolution of ALPR into true Vehicle Recognition. At the same time, PlateSmart has had great success in advancing video analytics through its groundbreaking technology. The idea is not to dictate to the industry what must be done, but to show what is possible and as PlateSmart CEO, John Chigos, believes what is inevitable. The ability to identify vehicles using multiple factors besides just license plate numbers, combined with the ability to analyze this data and produce true actionable intelligence in real time, is the undeniable future as more and more cities adopt citywide surveillance.

The ability to identify vehicles using multiple factors besides just license plate numbers, combined with the ability to analyze this data and produce true actionable intelligence in real time, is the undeniable future as more and more cities adopt citywide surveillance.

John Chigos and his team have created a suite of solutions that not only allow unprecedented identification and tracking of suspect vehicles, but also employ the latest safeguards to protect personal privacy. Privacy is one of many core values for PlateSmart. John Chigos’ views have received support from a privacy advocate, Electronic Frontiers Foundation. Simply put, PlateSmart is showing that the law can be enforced and the public protected without violating anyone’s civil liberties. As an added benefit, the technology removes any question of law enforcement profiling as it only analyzes data and does not discriminate based on any factors other than vehicle identifiers and vehicle behavior.

PlateSmart has also demonstrated its forward thinking on the issue of data security. A cornerstone of PlateSmart’s founding philosophy is that vehicle capture data should remain strictly in the hands of law enforcement and security agencies, and should be freely shareable among these organizations without having to purchase it. To this end, the PlateSmart Network has been developed to allow this secure sharing and eventually enable statewide fusion centers to share with other agencies nationwide. By eliminating third-party data storage, PlateSmart aims to keep data where it belongs while maintaining the integrity of the data.

All of the goals discussed here are achievable. It is incumbent upon us in the security industry to work with lawmakers and agencies, and in so doing, help to create a safer world.

In conclusion as industry manufactures we advocate industry players to work together towards the same goal, a safer and more secure country. By adhering to ALPR best practices, using advanced technology such as this, reduces concerns of maintaining data, data wholesalers providing old and inaccurate information, eliminates concerns of profiling, since the camera can only analyze vehicle identifiers and behavior, without intruding on privacy rights.