Cold Starts and Searching Part 2
By: Jerry Bradshaw
Police K9 Magazine
In part 1 of this two part series of articles we discussed proofing the “alert” and suspect encounters in the building search as well as area searches. We discussed issues related to finding the subject out in the open vs. in hiding and many associated exercises that make building searching and area searching more realistic so that there is not a drop in performance between training and a real deployment. In this second part of the article, we continue to work backwards in the training sequence from the alert and encounters with the suspect, which we discussed in part 1, to the actual hunting for the man and the start of the search. Our goal is to highlight important issues in proofing the hunting behavior and the start sequence so that there is no decline in performance between training and deployment.
Just like in detection, building and area searches in controlled aggression require a dog with strong hunting drive. The overriding value of a police dog is in his function as a locating tool. The K9 is much more efficient in locating hidden suspects in the dark and that efficiency extends to the benefit of allowing for increased officer safety in these high risk encounters. How much a dog has to hunt and how we teach the dog to hunt properly is an extremely important feature of doing building searches. Hunting properly for a suspect who is potentially violent requires teaching the dog to be precise in his hunting behavior so that he doesn’t clear an area that really isn’t clear.
Many times I have observed trainers placing the decoy in a building for a building search deep in the back of the building. They do this because they want to “exercise” the dog’s hunting. When this becomes the norm for the dog in training, the result of this failure to move the decoy around in proper hide placement sequences is to teach the dog to race into the building and bypass doors and thresholds with barely a notice until he gets well deep into the building. Only then does the dog start searching in earnest for the subject.
Imagine doing a detection exercise on a vehicle exterior and you only put the hides (and consequently the reward) at the rear of every vehicle. Over time the dog will learn that searching the front of the car, even if you present it, is not profitable and he will blow past the start of the search in favor of getting to the area in the vehicle where he finds the contraband and the resulting reward for which he is working. We would never be so foolish in detection problems. In detection we make hide placements in varying productive areas. Sometimes the hide placement is right at the start of the search sequence to make the initial presentation meaningful to the dog (he finds things there just as likely as deeper into the search pattern). This makes the dog want to explore all the seams in the car as equally likely to produce target odor and consequently a reward. Extending this to a building search means we must place the decoy close to the insertion point and vary the searching required so that all hide placements from shallow to medium to deep are equally likely.
Many in service training days give each team only one repetition of the building or area search and this is a huge mistake. Dogs need meaningful repetitions to learn and so you must do more than one repetition in a train- ing session. This is why I like to choose buildings that have multiple points of entry. You can hide the decoy in the same general area, so that you do not create multiple scent pools, but vary the insertion point to create short, medium and longer hunts that you can vary. You get the further benefit of the dog seeing varying points of entry and so the “building” looks different each time.
For example, I might set up a building search scenario where the dog is inserted and finds the decoy as soon as he crosses the threshold of the entry point, around the corner (a visual find); the next one might be slightly deeper (with decoy hidden); the third one might be even deeper than the second entry but with the decoy in the open; and the last one closer to the starting point and hidden. This process of varying the depth of finds in the hunting will produce a dog that hunts methodically regardless of the depth in
the building. As I explained in part 1 of this series, I will always vary the decoy’s placement between being hidden and out in the open. When you vary the hide placement, the dog must give his alert when the decoy is hidden, go for an immediate apprehension when the decoy is in the open or have to use effort to get to the decoy when partially obscured.
Additional variables in hunting
As the dog progresses in the hunting we will also vary the height of the suspect decoy and the depth away from closed doors when he is hiding and proof the dog off of equipment by leaving dead equipment out in the open and behind closed doors so the dog learns the equipment is not the predictor of his reward, but rather human odor. In area searches, moving the suspect decoy high in trees and also low under tarps covered with pine straw or other brush is critical. As I mentioned in Part 1, dogs must be exposed to dead equipment during obedience training drills, so they learn it is not relevant to them. If the dog goes in and self-rewards on dead equipment, that is your signal to work on the issue in obedience first and then make it so the dead equipment is behind doors and inaccessible so that you can extinct the behavior of the dog for going to it and refocus him on hunting for the man. If your dog has a strong interest in dead equipment, place the equipment in a wire dog crate in the search area so it is not possible for him to bite it and you can correct his interest in it and refocus him on hunting.
Another variable that must be addressed is what kind of building or area the dog is hunting. I mentioned this issue in part 1 but it bears repeating that the dog must be fluent in hunting both institutional buildings with opposing doors off of a main hallway, more residential structures, barns, garages, sheds, car lots and areas that might have a mixture of structures like barns, out buildings, boats and parked cars. Practice in low light and nighttime conditions as well.
One of the most critical issues we need to proof in hunting is the ability to make a find and then continue to search after an apprehension. In training we have to stop the tendency to be “one and done.” That is, during most building search or area search train- ing, the set up calls for one decoy hidden in an area so that when the dog locates the decoy, the dog is finished and removed from the search area. The reality is nobody commits crimes alone and if you found one suspect, you likely have to locate another. Imagine you are doing a building search and your dog indicates on a closed door and you reward the dog by having the decoy pop open the door and give a grip. The dog is then removed from the bite, decoy is placed in custody and is walked out of the building. The dog must be able to be refocused on searching after finding the first suspect decoy. Many dogs struggle with this at first. Therefore, we would place a second decoy close to where we start the resumed search so the dog can quickly locate and find a second suspect decoy and be rewarded quickly for switching back into a hunting mindset. We hook up and walk out decoy number 2. Now there is a 3rd decoy further into the search area. You can see the point. We then have to vary the hunting difficulty on the second and third re-deployments to match the dog’s capability.
Quick rewards in the beginning show the dog that there is more to be had by going forward rather than obsessing over the one who was marched out. When the dog can
do 2-3 complex searches in sequence, you can consider yourself in the minority of K9 teams that can really search a building competently. Each find could have different equipment profiles. One could be in civilian clothes only and is an alert on a closed door, recall to cover and call the suspect out to an apprehension with-out a bite. Another could be a low profile hidden sleeve reward for finding the subject on the second deployment of the search where the suspect is behind a doorframe, out in the open. The last one could be a decoy in a full bite suit behind a door, where there is a dead bite suit across the hall behind a door.
The Start: Hot, Warm and Cold
As this article comes to an end, we now are ready to discuss the start of the search process. In dog training we call this process back-chaining. Teach the dog how to deal with the suspect (last thing he will do in the sequence of a building or area search) by teaching proper biting behavior or alert behavior if the suspect decoy is hidden (barking or passive indication). Then add in the hunting, make it more and more complex, and finally we get to how we start the dog.
The start in a building search or area search, in my mind, is very much like the start in tracking. If you cannot roll up on a scene to a “cold” start in tracking (where you have only a point last seen and maybe a direction of travel) and cast the dog to locate the start of the track, cut the path and get on the track and move out, it doesn’t matter how well your dog can actually track once on the track. Picking up the start is the most critical piece of the process.
In building and area searches we do what we call hot, warm and cold starts. In hot starts, the dog is learning that searching buildings and areas will result in an opportunity to bite. The dog is brought to the building and we allow him to see the decoy. We have the dog alert on the passive decoy with barking. The dog’s barks make the decoy come alive and run into the building for the dog to chase and consequently bite. After a few of those repetitions, the dog is posted at the doorway and we give his alert command and allow him to frustrate. Upon barking, a decoy already hidden inside, out of sight is “barked out” to the dog posted at the threshold. In an area search, we use the engine block of the K9 vehicle as the anchor point of the search, so the dog comes to understand this point of cover is from where he will start his area searches. This is the beginning of warm searches. From these “bark-outs” we start removing the need for visual stimulation to cue the dog that he is searching for a hidden man. He learns that if he activates in drive the man will appear.
Alternatively, we can start with a bark out and the decoy comes half-way to the dog and then runs away into the building or search area and hides. The dog is released to a simple search where the dog hunts quickly and finds the subject out in the open, for a bite reward for a simple hunt. These progress quickly to a bark out and instead of the dog seeing the decoy run away (visual prey stimulation) we pull the dog back from the doorway where he cannot see the decoy go hide and give the dog 15-20 seconds of “break contact” time and then repost in the exact same spot and send the dog in for his hunt and consequent bite reward. From this, the dog learns that he doesn’t have to see the decoy to go hunt for him.
These warm searches are the most critical. The dog is gradually increased in time from an initial break-contact time of 15-20 seconds to a minute to 2 minutes, to put- ting him back in the car for 5 then 10 and up to about 20 minutes. The dog is always returned to the exact spot from where he made the visual contact in the bark out but now the decoy is hidden. Keep the search patterns simple at first as you vary the time of the warm starts. As the dog gets more con dent in his starts after more minutes of breaking contact, you can increase the difficulty of the searching.
Finally we introduce cold starts. I normally recommend going back to the same building insertion point you did one or two days prior with warm starts. Use some of the dog’s residual memory of the event. Here we bring the dog to the threshold and start the search without any visual stimulation. The hunt itself is super simple, with the decoy just around the corner of the insertion point. After that quick reward, vary the insertion point to make the hunt slightly more complex. Return the dog to the car and after he has been in the car for 30 or so minutes, drive around a little and roll up to the next insertion point on the building and send the dog in to search. You should see when you post up and hold his collar or allow the dog to pull into the line wanting to go forward into the building. If the dog seems confused or not coming up in drive, go back to warmer starts. The habit is not yet set. If the dog takes to the cold starts well, variably increase the complexity of the hunting.
As we did with hunting itself, vary the look of the start from entry ways of different buildings, rooms, bottom of stairs cases in barns and finally attic insertions for buildings. For area searches, the engine block of your patrol vehicle parked at the edge of a wood line, field, car lot and anything else you can think of where you might do an area search is imperative. Until you have generalized this behavior in all these environments, you cannot be sure that the dog will understand what you want of him when faced with a new building or area. When changing areas around, go back to short hunts and quick rewards, then make the hunting more complex and the suspect encounter more difficult (e.g. the dog searches a car lot and the suspect decoy is inside a closed car). Do multiple repetitions within each session.
During the hot start phase of training, the dogs can hunt longer and perform more complex searches for the simple reason that they are starting off in a very high drive state from the visual stimulation. As we add the complexity of removing the visual stimulation at the start, we simplify the hunting so the dog gets to the reward more quickly. This means it is important to make good inroads on the hunting work while the dog is still in the hot phase of the start. The alert on concealed suspects can be started completely away from any complex hunting, so that the alert is developed somewhat in isolation from hunting as we do with the final response in detection.
You will notice I haven’t said anything about announcements. I do not think it is tactically sound to give away your position from which you will insert the dog by giving an announcement from that position verbally. Case law and agency SOP require announcements. I prefer to have another officer do any announcements over a PA system to keep your deployment tactically sound. You may give your dog a whispered “announcement” if you wish to get him up in drive before sending him in or you can rely on the context of the deployment from a position of cover and concealment. If you deploy on a long line or in a special harness, that added context will help as long as you have been consistent throughout training with associating this equipment with searching. I want the dog to really understand the command word for searching for a human quarry, so be consistent with that in your building search training.
As you can see, there are many variables to consider in your training when it comes to area searches and building searches. Skipping the warm starts is one of the biggest mistakes that I see. Skipping from hot to cold starts for many dogs will create a lot of confusion. Often when trainers do this and see that the dog is confused, they have the decoy make noise to attract the dog’s attention. This is a huge mistake. The dog will come to rely on the decoy cueing him that in fact there is a search for a human suspect to be done.
This will lower the likelihood that the dog will be able to do a proper cold start. The warm starts successively approximate the cold start, which is our ultimate goal. Concentrate on changing one variable at a time and be aware of when you try to do too much in one session. When you arrive at in service training and are set to do another building search in the same building with the decoy in the same old place, have a sit down with the other handlers and discuss how you can proof all of the elements of the building and area searches! Happy hunting!