Are we working too clean?

Here’s the uncomfortable thought that keeps coming back to me after all these years working with detection dogs: I’m not sure we are preparing our dogs well enough for the real operational world. And maybe even more confronting… I sometimes feel that, as a K9 community, we are trying to keep everything a bit too clean.

Don’t get me wrong. I understand odor hygiene. I use gloves when I’m working with toxic substances. I start my dogs on clean, pure target odors when I imprint them. I clean my equipment carefully, using isopropanol, heat, and structured procedures. I know how fast dogs can pick up the wrong associations. Cross-contamination, overshadowing, unwanted pairing, these are real risks, and they matter.

During a workshop in the United States, I was confronted by someone who was genuinely upset that I handled target odor with my bare hands. For her, that was unacceptable. It became a discussion, not really about learning, but about being right. Later that day, I asked if I could use some of her training aids. She opened a plastic toolbox. Inside were multiple containers, bags, and different target odors all mixed together. No structure. No separation. Just one big odor soup. It was one of the clearest examples of cross-contamination you could imagine.

That moment stayed with me. Not because someone made a mistake, we all do, but because it showed something deeper. Sometimes we focus so much on one rule that we lose sight of the bigger picture.

When babies are born, they are fragile. But they don’t grow stronger in a sterile box. From the very beginning, they are exposed to bacteria. Through their mother, through their environment. As they grow, they play outside, get dirty, interact with the world. That exposure is not a problem, it’s part of building a strong immune system. If you protect them too much, keep everything too clean, their system doesn’t learn how to deal with reality.

I know how extreme the other side looks as well. As a cancer survivor, I’ve lived in a completely controlled, sterile environment. After chemotherapy, my immune system was almost gone. For a long period of time, everything had to be clean. People around me wore protective clothing, gloves, masks. That level of hygiene was necessary. But it was never the end goal. Step by step, I had to return to the real world. A world full of bacteria, viruses, unpredictability. Because staying in isolation would not make me stronger. Exposure, carefully managed, was what rebuilt my resilience. And that’s exactly where I see the parallel with detection dogs.

Criminals don’t use pure target odors. They don’t store things in perfect containers. Drugs are mixed to increase profit. Explosives are often homemade, using whatever materials are available. Targets are hidden deep, layered, masked with strong distracting odors, coffee, food, chemicals, waste. People are creative when they don’t want to be found. And then we take our dogs, trained on clean, controlled, predictable setups, and expect them to perform flawlessly in that environment. That’s where the gap starts to show.

I’ve seen many training programs where dogs work in the same few locations over and over again. The same buildings. The same people. The same routines. The dogs arrive and already know what’s going to happen. You can almost see it in their behavior: expectation. “This is the place where I find things.” And usually, they do. Because that’s how the training is designed. Even when we change locations, it’s often to empty buildings. Places where it doesn’t matter if a dog scratches something or jumps on furniture. But those environments are often odor-poor. They don’t reflect real life. There’s little complexity, little variation, little pressure.

Then suddenly, that same dog is deployed into a real home. A place where people live. Where there’s food, laundry, pets, cleaning products, daily life. A completely different odor picture. A completely different atmosphere. And the dog hesitates. Not because the dog is not capable, but because the situation doesn’t match what it has learned to expect.

Dogs don’t just learn odors. They learn environments. They learn patterns. They learn where success is likely to happen. And they build expectations around that. If we only train in predictable environments, we create predictable dogs. And the real world is anything but predictable. This is where I think we sometimes shoot ourselves in the foot. We aim for perfection in training, clean odors, controlled setups, structured routines, but in doing so, we risk creating dogs that are very precise… and at the same time, not flexible enough.

In research, you need control. You need to eliminate variables. You need clean odors, proper storage, controlled airflow, consistent conditions. That’s how we learn what actually influences behavior and detection. But operational work is not a laboratory. And that’s where we need to be careful. Because if we take that same level of control and try to apply it rigidly to real-world training, we may end up preparing dogs for a world that doesn’t exist.

Start clean. Build clarity. Teach the dog what the target odor is. But then… gradually let go of control. Introduce variation. Mix odors. Change environments. Train in places where people actually live and work. Let the dog experience distraction, complexity, uncertainty. Not in a chaotic, random way, but in a thoughtful, progressive way. Because that’s how you build resilience. That’s how you create a dog that doesn’t just perform under perfect conditions, but can adapt when things are messy, unpredictable, and unclear. And that’s what operational work demands.

This is not about being right or wrong. It’s not about criticizing people who work clean or people who work more “rough.” It’s about asking ourselves an honest question: Are we preparing our dogs for the world they will actually face? Or are we preparing them for the world we feel comfortable controlling?

The people who know me, know that I love science and that I follow it closely. And yes, in the beginning of a dog’s career, I train in a very controlled way. I use line-ups, scent wheels, sensors, data, and structured environments to measure and understand the performance of the dog before moving toward the operational side of training. Because I strongly believe that the foundation of a detection dog must be solid. There is no room for shortcuts. If you take them early on, they will catch up with you later, and often at the worst possible moment. And once you’re there, it’s not easy to go back and fix it.

I’ve seen too many operational dogs struggle with something as simple as a line up. I’ve seen dogs fail an odor recognition test because the source came from a different training kit. I’ve seen dogs perform well in training, but fall apart in real operations because the environment didn’t match their expectations. And that’s the point where it all comes together.

This is about thinking clearly, planning with intention, and then doing the work that truly prepares the dog. It connects directly to the principles behind the Matching Law, where behavior is shaped by the history of reinforcement and the conditions in which it was learned. If we train in clean, predictable environments, we reinforce behavior that fits that world. But if the operational world looks different, and it always does, then we should not be surprised when performance drops. So know what you are training. Understand what your end goal really looks like. And most importantly, prepare your dog for the real world. Because that world is not clean. It is not controlled. And it is definitely not as sterile as many of our training scenarios make it seem.