COVID-19 Biodetection Dogs at Airport Pilot Programs

The Pandemic-Era Pilot Programs

During 2020 and 2021, several airports and venues deployed trained dogs to screen passengers or attendees for possible SARS-CoV-2 infection. The Helsinki-Vantaa airport in Finland was one of the most publicly visible pilots, with passengers asked to provide a sweat sample swabbed from the neck or arm that was then evaluated by a trained detection dog. Beirut International Airport in Lebanon ran a similar pilot. The Miami Heat’s NBA basketball arena piloted detection dogs at venue entry during a portion of the 2020-2021 season.

These programs were screening adjuncts, not clinical diagnostic tests. The premise was that infected individuals — symptomatic and asymptomatic — produce VOC signatures in sweat and breath that trained dogs can detect, and that a positive dog indication could direct passengers to confirmatory PCR testing. The pilots ran during a period when fast turnaround PCR was scarce and rapid antigen testing was still being deployed; the dogs represented an attempt to add screening capacity quickly.

The pilots received significant press attention and produced peer-reviewed research that contributed to the broader biodetection literature. Most have wound down as the pandemic acute phase eased and other screening tools matured.

The Underlying Research

The most-cited published research on canine COVID-19 detection came out of a collaboration led by Grandjean and colleagues at the National Veterinary School of Alfort in France, published in PLoS ONE in 2020 and followed by additional studies. The research demonstrated that dogs trained on sweat samples from PCR-confirmed COVID-19 patients could discriminate between positive and negative samples at meaningful rates in controlled study conditions.

Reported sensitivity in some published studies has been above 90 percent for symptomatic samples, with somewhat lower numbers for asymptomatic samples in various studies. Specificity has also been reported high in many studies. These are encouraging research numbers. They came with caveats — small sample sizes in some studies, single-site rather than multi-site validation, and questions about how the methodology would scale.

Medical Detection Dogs in the UK contributed further research, as did Penn Vet Working Dog Center, the National University of Colombia, and several other research programs. The body of literature on canine COVID-19 detection grew substantially during 2020 to 2022. The general finding across studies was that the proof of concept holds — trained dogs can detect SARS-CoV-2 infection from sweat or breath samples in controlled conditions — but the operational translation challenges are substantial.

What the Pilot Programs Actually Did

The Helsinki-Vantaa pilot, which ran from late 2020 into 2021, was probably the most operationally mature. Passengers volunteered to participate, provided a sweat sample on a gauze pad swabbed from the neck or arm, and the sample was evaluated by a trained detection dog in an adjacent screening room. Positive indications referred the passenger to confirmatory PCR testing. The pilot also collected research data alongside the operational screening, contributing to the published evidence base.

The Beirut pilot followed similar protocols. The Miami NBA pilot was venue-based rather than border-based — attendees provided a sample at entry, dogs evaluated samples, and positive indications referred to confirmatory testing.

The honest assessment of these pilots is that they were screening adjuncts that supplemented PCR and antigen testing rather than replacing either. No serious public health authority claimed the dogs replaced laboratory testing. The pilots demonstrated that screening could be added quickly with a different sensor modality, contributing to the layered defense against pandemic spread.

The Limitations Honestly Stated

Canine COVID-19 biodetection had real limitations that the research community has been transparent about. Sample sizes in many studies were small, limiting the precision of sensitivity and specificity estimates. Asymptomatic detection rates were often lower than symptomatic detection rates, and asymptomatic transmission was a major epidemiological driver. Variant-specific performance was uncertain as the virus evolved through Alpha, Delta, and Omicron waves.

Training dogs on COVID-positive samples and maintaining the training as the virus evolved required sustained sample sourcing, which became harder as pandemic phase eased and PCR-confirmed positive samples became less abundantly available for training. Operational deployment in busy airport environments required handler-team training, throughput logistics, and integration with confirmatory testing that limited scalability.

The honest framing is that the dogs were never a replacement for PCR or antigen testing, were not framed that way by serious public health programs, and remain a research-and-pilot capability rather than a standard screening tool.

The Broader Biodetection Context

Canine COVID-19 detection sits in a broader infectious disease biodetection research family. Medical Detection Dogs in the UK has published on canine detection of malaria from worn socks, tuberculosis from sputum samples, and several bacterial infections. APOPO, a Belgian-founded nonprofit, has famously trained African giant pouched rats for landmine detection and for tuberculosis detection — a different species but a parallel biodetection mission.

The same VOC-detection logic underlies cancer biodetection research, where trained dogs detect cancer-associated volatile compounds in patient samples. Both fields share the proof-of-concept-strong, clinical-translation-hard narrative. Both contribute methodology and findings to electronic nose and biosensor development. Both face similar regulatory and scalability barriers.

The training methodology overlaps with conservation detection covered in Working Dogs for Conservation scat detection and pest detection covered in bedbug detection dog NESDCA certification. The underlying ability of dogs to discriminate complex VOC signatures is a common thread across many detection disciplines.

Where the Work Has Gone Since

Most COVID-19 detection dog pilots have wound down as the pandemic acute phase eased, fast turnaround testing became widely available, and the layered public health response shifted. Some programs have pivoted to other infectious disease biodetection research — using the trained dogs and infrastructure to study detection of other pathogens or to support continued biodetection methodology work.

The research findings remain meaningful as proof of concept that canine biodetection of viral infection is feasible. The future deployment context, if a future pandemic situation calls for rapid screening adjunct capability, could revisit canine biodetection as one tool among many. The infrastructure built during 2020-2022 has not been entirely dismantled in most programs, and the research community continues to publish.

What This Means for Public Health Planning

Canine biodetection should not be central to pandemic preparedness planning, but it should be on the list of available rapid-deployment screening modalities for situations where conventional testing is scarce, slow, or insufficient. The pilots demonstrated that the capability can be deployed in weeks to months — much faster than developing and scaling a new molecular test.

For routine public health response, evidence-based laboratory testing remains the standard. Canine detection is best understood as a niche capability with proven proof-of-concept that might play a layered screening role in specific scenarios. It is not a replacement for the broader public health infrastructure.

How the Dogs Were Trained

Training for COVID-19 detection followed the general detection-dog methodology with COVID-specific sample sourcing. Dogs were imprinted on sweat samples from PCR-confirmed positive patients, proofed against samples from PCR-negative controls, and validated on held-out samples before deployment. Double-blind protocols were essential to credible findings — without them, the Clever Hans risk (dogs cueing off handler subtle cues rather than the sample) would undermine the work.

Sample sourcing happened under research ethics protocols with patient consent. Some samples came from hospitalized patients; some from outpatient screening centers; some from research-volunteer protocols. The methodology was published in detail in the peer-reviewed literature, supporting replication and meta-analysis as the field has matured.

Welfare in Biodetection Work

Biodetection dog welfare during pandemic-era deployment was a meaningful concern. Working in busy airport environments during a pandemic raised questions about handler exposure risk, dog stress in high-foot-traffic environments, and the sustainability of high-tempo screening shifts. The best programs built in rest, decompression, and shift rotation that mirrored the broader working dog welfare standards framework.

The dogs themselves did not appear at high risk of contracting SARS-CoV-2 from screening work in any documented program. Some research showed dogs could be infected experimentally, but operational screening exposure did not produce documented infections in working dogs in the published literature.

Where Biodetection Goes Next

The honest future of canine infectious disease biodetection is probably as one tool in a broader biodetection toolkit that increasingly includes electronic nose technologies, point-of-care molecular tests, and other rapid sensor modalities. The dogs may continue to play a role in proof-of-concept research, in specific operational niches, and in informing the design of next-generation sensor technologies.

The work is interesting and the field has produced credible peer-reviewed research. It has not, and probably will not, replace laboratory-based diagnostics for infectious disease. That honest framing should ground any discussion of where biodetection dogs fit in public health planning.

Frequently Asked Questions

Did COVID-19 detection dogs replace PCR testing?

No. The pilots were screening adjuncts that referred positive dog indications to confirmatory PCR testing. No serious public health program framed the dogs as a replacement for laboratory testing. PCR remained the diagnostic standard throughout the pandemic.

Are COVID-19 detection dog pilots still running?

Most have wound down. Some research programs continue to publish and some pilot infrastructure has pivoted to other infectious disease biodetection research. The acute pandemic deployment context that drove the original pilots has eased.

What sensitivity did the dogs achieve?

Reported sensitivities in published studies have ranged widely depending on study design, sample type, and patient symptom status. Some studies reported sensitivities above 90 percent for symptomatic samples in controlled conditions. Asymptomatic detection has been lower in many studies. These are research-context numbers, not clinical performance in routine screening.

Can dogs detect viruses other than SARS-CoV-2?

Research has examined canine detection of malaria, tuberculosis, several bacterial infections, and other viral pathogens. The proof of concept for canine infectious disease detection extends beyond COVID-19, but each pathogen requires its own training and validation work.

Can a dog tell if I have COVID-19 today?

No, not as a clinical service. If you have symptoms or exposure concerns, evidence-based options are rapid antigen testing at home or PCR through a clinical lab. Canine biodetection is not clinically available as a diagnostic for individual patients.