Dallas police wireless cameras tackle crime: an interview with Lt. Tony Crawford

Last week I interviewed Lt. Tony Crawford of the Dallas Police Department to get a better picture of how the they are using wireless cameras installed in various parts of the central business district (CBD) to tackle crime. I wanted to know more about this project after reading a newspaper article on how the cameras captured a robbery at a Dallas bus station and allowed the police to arrest the suspect within a short period of time.

Among the questions I posed to Lt. Crawford:

  • How many cameras are installed and what is the area of coverage?
  • Did the images captured on video help the police department track down and arrest suspects?
  • How do they measure “success”?
  • Was there any opposition to the project? Who was skeptical?
  • What were the most serious technical issues they encountered?
  • What precautions have they taken to assuage the fear of invasion of privacy among local residents?

There’s a lot of interest right now in surveillance via wireless cameras or old-fashioned CCTV which is popular in the UK. An article in today’s Guardian, a UK newspaper, says that UK police believe CCTV has been a “failure” because despite having spent billions of pounds on the cameras, they have led to only 3% of crimes being solved by CCTV. Of course, you say, people — not cameras — solve crimes. Indeed the Dallas police know that all too well and have designed their project accordingly.

Scope of the project

There are 40 wireless cameras installed in the Dallas CBD in areas recommended by police officers who regularly walk or patrol the CBD. 32 of the cameras are pan-tilt-zoom, that is, their movement is controlled remotely. The remaining 8 are pointed in one direction. They hang 20 feet off the ground in a white box which is the size of a bread box. The police department’s logo is on the box. They get electric power from the lamp posts. The cameras can be moved around and installed in other locations, for example, to monitor the area around a recently opened homeless shelter near Dallas city hall. The cameras, which are supplied by Bearcom, communicate with wireless nodes from Firetide; the wireless network uses the 4.9 GHz frequencies which is reserved for public safety.

There are two monitoring stations that are staffed 24 hours a day. The department has plans to install more cameras in the CBD and to add 2 more monitoring stations. As you will see in the last part of this article about the UK CCTV fiasco, intensive monitoring of images is critical to the success of any surveillance project.

Success in finding suspects and the measure of success

Two months ago, the department’s cameras helped in the arrest of a person on the FBI’s most wanted list (a camera near the library caught his image on video). More recently, police arrested a homeless man who was caught on video robbing another man in a Greyhound bus terminal. While the mainstream media equate “success” with these sensational stories, local residents and businesses are just as concerned with the smaller incidents of theft from parking pay boxes and with capturing images near reported crime scenes (thefts from homes and businesses). So far the police have arrested 317 people based on the video images and respond to numerous calls concerning various incidents in the area.

Crime is down 12% in the CBD but Lt. Crawford points out that one can’t say it’s all because of the cameras. The cameras are only one component in the recent drive to improve policing in the area and while they help the police, they cannot replace them.

Opposition, skepticism

I was not surprised to hear from Lt. Crawford that much of the skepticism and opposition to the wireless camera project came from the police officers themselves. All of us, police officers included, are wedded to our old ways of doing things. As in enterprises where the boss brings in a new computer system and the employees groan, so it is in police departments and perhaps even more so there, because the work that they do is dangerous. Structure is very much a part of police officers’ working lives and their habits are difficult to break. So the cameras had to prove themselves first before the officers could be convinced that they were tools that help in police work.

Fear of invasion of privacy

Outside the police department, residents and civil liberties groups expressed concern about the invasion of privacy. What is captured on video? How long are those images kept? How are they used?

The images are kept for 14 days. The cameras do not pan up to the loft and apartments so they do not look into the windows of residents. Any potential evidence captured on video is sent to an investigator.

Serious technical issues

The biggest issue with the cameras is cleaning the lenses. After a rainstorm, the lenses are caked with dirt and the service contract with Bearcom does not include cleaning. As a result, the municipality (with the permission of Bearcom) gets the street cleaning crew to do that.

Another issue is the quality of the video. The cameras read license plates from two blocks away and can make out faces one block away, depending on the light (less so at dusk). If it’s windy and the poles are shaking, the video images will be blurred.

Comparison with UK CCTV debacle

Today I read on the Guardian’s website that the CCTV (cameras) installed in so many cities in the UK have been a failure in dealing with crime. Here’s an excerpt from the article:

“Use of CCTV images for court evidence has so far been very poor, according to Detective Chief Inspector Mick Neville, the officer in charge of the Metropolitan police unit. “CCTV was originally seen as a preventative measure,” Neville told the Security Document World Conference in London. “Billions of pounds have been spent on kit, but no thought has gone into how the police are going to use the images and how they will be used in court. It’s been an utter fiasco: only 3% of crimes were solved by CCTV. There’s no fear of CCTV. Why don’t people fear it? [They think] the cameras are not working.” More training was needed for officers, he said. Often they do not want to find CCTV images “because it’s hard work”. Sometimes the police did not bother inquiring beyond local councils to find out whether CCTV cameras monitored a particular street incident.”

The problem is that technology in this case has been used to replace, not enhance, police work. It’s a “cheap” solution that in the end, is not so cheap after all, like companies that replace good technical support for cheap and not so good technical support.

Lt. Crawford told me last week, when I asked if they were going to add more cameras, that the department does not want to do so unless they can add more monitoring stations. He says that cameras will not be effective in helping the police if there don’t have enough people looking at the images, picking out the ones that are potential evidence.

This is the mistake that UK has made. They do not allocate enough resources to the police to analyse the video material. For the councils in cities, it’s easy to give people a false sense of security by putting up CCTV. But someone has to monitor and analyse the images. Indeed, in London, the police have decided to go a step further (excerpt from Guardian article) after the CCTV debacle:

The Viido unit is beginning to establish a London-wide database of images of suspects that are cross-referenced by written descriptions. Interest in the technology has been enhanced by recent police work, in which officers back-tracked through video tapes to pick out terrorist suspects. In districts where the Viido scheme is working, CCTV is now helping police in 15-20% of street robberies. “We are [beginning] to collate images from across London,” Neville said. “This has got to be balanced against any Big Brother concerns, with safeguards. The images are from thefts, robberies and more serious crimes. Possibly the [database] could be national in future.”

Click here to listen to the interview with the Metropolitan Police.


While technology (from wireless cameras to access points and wireless devices) can improve municipal operations and services dramatically, you can never take the human out of it. Time and time again, people have made the mistake of replacing humans with machines — in some cases it works well, in others, it’s a disaster. This applies not only to government services, but also to technical and phone support in enterprises. Who has not had the miserable experience of going through endless loops of automated telephone “support” (the worst offenders being the airlines and insurance companies)? The dismal failure of CCTV projects in the UK comes not from camera malfunctions but from lack of humans analysing the images. It’s not enough to say, well, you get what you pay for. In the UK example, they paid billions of pounds. It’s just as important to allocate the money effectively, both for the purchase and maintenance of the cameras, and for the employment of humans to analyse the images.


  1. Mick Neville says

    I am the DCI mentioned in your article – please can you send me the contact details of Lt Crawford. The article is very balanced – people are more important than equipment! My systems to make an end-to-end process have now been introduced & we have now arrested 500 criminals using CCTV.