Six years ago, an internal report commissioned by the Bureau of Alcohol, Tobacco, Firearms and Explosives issued a dire warning: The agency needed to open a long-standing, sole-source ballistics contract with a company based in Canada to competitors, or risk losing one of its most important tools for preventing and solving gun crimes. 

The report — written by an independent consulting company and provided to The Trace by a source who had access to it through work but asked to remain anonymous — said competition was likely to bring down prices, which would enable many police departments to access the National Integrated Ballistic Information Network, or NIBIN, for the first time. It might also spur innovation, since a sole-source vendor has less incentive to improve than a company that is competing to win business. 

But in 2019, the Bureau of Alcohol, Tobacco, and Firearms (ATF) renewed its $85 million contract with Ultra Electronics Forensic Technology for five years — a decision that experts say has likely impeded law enforcement’s ability to investigate shootings at a time when gun violence is surging. 

It is unclear why the single-source contract has endured. But several former ATF officials, including the agency’s acting director of four years, have gone to work for Ultra Electronics. While they appear to have followed the letter of the law when it comes to accepting employment with government contractors, experts said it still raises concern that cronyism could have colored the agency’s decision-making. This could give the “appearance of selling government contracts for personal gain,” said Craig Holman, a governmental ethics and lobbying expert for the nonprofit consumer advocacy organization Public Citizen.

Asked why the advice and timeline outlined in the report were not followed, an ATF spokesperson said in an email: “The evolution of the program has significantly changed from 2015. The program continues to assess and shift based on the current climate and priorities of the future and current NIBIN partner sites.” 

As to whether there are plans to open the contract to competition in the future, the spokesperson wrote, “ATF adheres to the guidelines … and awards requirements …  with respect to all vendor contracts, including contracts for NIBIN equipment and services.”

Stacy Stern, vice president of sales and marketing at Ultra Electronics, said her company would cooperate fully with efforts to open the contract to competition. “Our mandate is to solve crime,” she said. “We don’t want to be a roadblock or an obstacle.” 

In the early 1990s, Ultra Electronics Forensic Technology became one of the first companies to develop a way to collect and compare images of bullets and shell casings, which can then be used to solve gun crimes. The company, which is based in Montreal, and was acquired by a British company in 2014, may soon be acquired by an American private equity firm. At least three other companies have developed similar technology, but they haven’t been able to gain a foothold, in part because images generated by their machines cannot be compared to the 4.5 million images already in the NIBIN network, which are proprietary and not available for public inspection. 

In its 2015 report, Interact Business Group said that NIBIN was “plagued with several structural deficiencies,” including “over-reliance on a single-source technology supplier,” inadequate training, and dependence on police agencies to do their own comparisons. At the time, the ATF had removed NIBIN equipment from dozens of agencies that it said were not using the technology efficiently, and only about 150 police departments were using NIBIN. Of those, only about 30 were considered to be even moderately successful by NIBIN leadership, the report said. 

If things didn’t change, the consultant concluded, “it is likely that NIBIN will suffer significant, if not complete, collapse.” 

In the years that followed, the ATF made several of the changes urged in the report, including improving training and communication with police agencies and creating a centralized office where staff could do comparisons for agencies across the country. But the sole-source contract with Ultra Electronics Forensic Technology has remained. 

ATF issued a “request for information” for other vendors before entering into the 2019 contract with Ultra Electronics. But two people with knowledge of the bidding process told The Trace that it would have been extremely difficult for a company to meet the criteria for the contract without access to the proprietary database. For example, the bid said that it was “highly desirable” for the company to be able to compare newly captured images with preexisting images — a capability that would be impossible to develop without access to the current NIBIN data.

An independent researcher with knowledge of ballistics-imaging systems and NIBIN said that sticking with just one vendor is harmful. The researcher asked to remain anonymous because their work regularly brings them into contact with the ATF. “Locking the database of digitized ballistics images, and giving the only key to the sole vendor, impedes innovation,” the researcher said. 

Thomas Brandon, who was acting director of the ATF at the time the report was issued, retired in 2019 and, about a year later, went to work for Ultra Electronics Forensic Technology, according to his LinkedIn page. Brandon, who is one of several high-ranking ATF officials to join the staff of Ultra Electronics before and after the 2015 report, did not respond to an interview request relayed through an Ultra Electronics spokesperson and by personal email.  

Neil Gordon, a senior researcher specializing in government contracts at the Project on Government Oversight, a nonpartisan independent watchdog that investigates and exposes government waste, corruption, and abuse of power, said the endurance of Ultra Electronics’ single-source contract seems to follow the letter of the law — but not its spirit. Such a contract, he said, “… can devolve into a monopolistic situation where the government has to rely on contractors solely and puts the government in a vulnerable position.” 

“It might result in a worse deal for taxpayers,” because prices and service are not competitive, Gordon added.

Brandon is not alone among NIBIN officials to later be hired by Ultra Electronics. James Needles, once chief of the ATF’s Firearms Operations Division, is now forensic intelligence strategy manager for Ultra Electronics. Pete Gagliardi, who helped establish the NIBIN program for the ATF in the 1990s, became Ultra Electronics’ senior vice president after he retired in 1999, according to his LinkedIn page. (Gagliardi has since left and become an independent consultant.) 

At least two other ATF officials who left the agency after the report was produced worked for the company as contractors after they retired. 

“That looks very bad,” Gordon said. “It might raise in people’s eyes the specter that [ATF] officials are trying to do favors for this company in hopes that they will be hired — that they are looking out for the best interest of the company, and not the public.” (At least two of the companies hoping to compete for the NIBIN contract also employ at least one former ATF official.)

Reached for comment, Gagliardi said that he followed all the ethics protocols, both internally and externally, when he went to work for Ultra Electronics Forensic Technology. The other former ATF officials either declined or did not respond to requests for comment. Several former ATF and NIBIN officials did say that they don’t believe Brandon or anyone else at the ATF favored Ultra Electronics for the promise of a future job. Rather, they said it was simple inertia: It was easier to maintain the current contract and deal only with one company, even at the possible cost of a better program. 

Some also said that the NIBIN program has been a point of pride for ATF, and maintaining the sole-source contract has allowed the ATF to retain control over it. 

NIBIN examines images of the unique markings that a gun leaves on shell casings when it is fired and compares them to determine whether casings were shot from the same gun. This is a huge asset when, for example, police collect shell casings at one crime scene, run them through the NIBIN database, and learn that they were shot from the same gun that was used in a different shooting across town. Connecting two shootings can help crack a case, especially when police have strong evidence — surveillance video of the shooter, an eyewitness, a license plate — connected to one of the crimes. In its 24-year history, NIBIN has been used to connect shell casings from different crime scenes to one another more than 300,000 times.  

But NIBIN is only as good as the information it contains. Ideally, police departments in small towns and big cities alike need to collect shell casings whenever a gun is fired — not just when someone is shot or killed — and enter the casings into the NIBIN database. Studies have shown that many police departments that have NIBIN are inconsistent when it comes to collecting shell casings and entering them into the database. And only about 250 of the country’s 18,000 police departments have NIBIN equipment. Smaller departments may take their casings to regional labs, which are often backlogged and can take so long to turn around results that the shooter has fled or been caught before police get any helpful information. 

Thomas Brandon, then the Acting Director of the Bureau of Alcohol, Tobacco, Firearms, and Explosives, testifies during a Senate Judiciary Committee hearing in 2017. AP/Carolyn Kaster

Bruce Houlihan, director of the Orange County Crime Laboratory, in Southern California, said his lab turned to EvoFinder — an Ultra Electronics competitor — in 2012, after ATF took back NIBIN machines that it had loaned to his lab. The lab could have bought its own NIBIN machines from Ultra Electronics, but Houlihan said Orange County wanted to test bullets as well as casings, and Ultra Electronics was going to charge twice as much. (A spokesperson at Ultra Electronics said similar equipment costs less today, though the prices she cited are still higher than those offered by other companies.)

EvoFinder, which started in Russia and now has offices in several countries, can’t communicate with the wider NIBIN network. While Houlihan said that such broader communication would be preferable, the EvoFinder system does connect shootings within Orange County. That’s the vast majority of the hits Houlihan gets anyway, so the local database has value.  

After seven years of using EvoFinder, Houlihan said, his lab has gotten 700 hits — a number he is pleased with. He said it would be better if police agencies and labs could choose whatever technology suits their needs and budgets, while all connecting to one central network. “Ideally, we would all share our data without worrying about what technology we’re all using,” Houlihan said. “A little competition in vendor space is a good thing because they all build on each other and keep improving.”  

Kenneth Novak, a criminal justice professor at the University of Missouri-Kansas City, said that if ballistics-imaging machines were cheaper or more portable, police would be more likely to submit shell casings to NIBIN, and that could help solve crimes, especially in big cities where there are a lot of shootings. But Novak also said other key factors would have to fall into place for NIBIN to reach its full potential. Police need to commit to picking up shell casings every time, not just when someone’s been shot or killed, and entering them into the system. In other words, getting more NIBIN machines to police would help, but there are also other factors at play. 

“The human factor is key,” he said. 

Stern, the Ultra Electronics spokesperson, said the machine that takes images of the shell casings costs about $140,000, and the one that compares them to the database is about $60,000 — or about $200,000 total. That’s roughly the average annual cost of two police officers, according to one national survey. Smaller departments can also lease the equipment for a “subscription fee” of $15,000 to $30,000 a year, Stern said. The 2015 report said that allowing companies to compete to provide police departments would almost certainly bring the price down — maybe dramatically. Experts interviewed by The Trace agreed that less expensive machines would be in reach for more departments. 

One potential competitor to Ultra Electronics, Illinois-based Ballistics IQ, has developed a smaller, cheaper shell-casing scanner that creates less detailed images than NIBIN machines, but can be toted around in a police cruiser. And experts said that increasing officer access to the machines could improve how many shell casings are added to the database, which would then expand the potential for crime solving. They compared it to the Integrated Automated Fingerprint Identification System — the international system used to collect and compare fingerprints. The FBI and the National Institute of Standards and Technology (NIST), a government agency charged with promoting innovation and competitiveness, set quality standards for fingerprint collection decades ago, then opened the contract to competition. Today, the technology is affordable — a basic machine can be purchased for as little as a few hundred dollars — and it’s accessible to just about any police precinct, courthouse, or jail that wants one.  

The 2015 report suggested that the ATF work with the NIST to come up with base standards for ballistic images. The report also suggested that the agencies pursue an open-source algorithm to correlate ballistic images. 

The recommendations were supposed to take about two-and-a-half years, with a goal of opening NIBIN to competition by 2018. While that has not happened, Johannes Soons, a scientist working on ballistics for NIST, said his group has partnered with ATF and Ultra Electronics to study the technical feasibility of making ballistic images that are sort of like consumer-facing JPG files, which can be opened and viewed by any application, no matter what brand of camera took the photo or what kind of computer is used to open it. Soons said it will be up to the ATF to decide what to do with the information. He would not comment on why the steps outlined in the study were delayed, but said “there were several issues that caused delays in this large and complex project.”

The researcher with knowledge of ballistics imaging and NIBIN said an obvious solution would be for the ATF to make a sample of images in the existing database available to prospective vendors, so they could work on making their images compatible. 

David Chipman, the Biden administration’s former nominee to lead the ATF, said that, if he had been confirmed, helping NIBIN live up to its potential would have been one of his priorities. That would mean ensuring current contracts were open to competition, as well as finding ways to incorporate other cutting-edge investigative techniques, like gunshot detection, DNA, and microstamping. 

“Historical barriers such as cost, ease of use, and timeliness of results must be overcome; either by current tech vendors or new competition,” he said. “A suite of old and new technologies that can actually identify the shooter must be the aim of NIBIN today.”