Grundherr Seminar

WILLY GRUNDHERR SEMINAR
Co-Sponsored by FEMA Nebraska Task Force 1 & FEMA Missouri Task Force 1
Lincoln, Nebraska
Jun 5-7, 1998

Over the weekend of June 5-7, 1998, I was privileged to attend a seminar given by Willy Grundherr and his assistant, Kathy Jo Carlisle. This seminar was co-sponsored by the FEMA Nebraska Task Force 1 ("NE TF 1") and the FEMA Missouri Task Force 1 ("MO TF 1") and was held in Lincoln, Nebraska. The focus of the seminar was "Teaching the Bark Alert" and "Building Drive in the Disaster SAR-Dog."

In a word, the seminar was fantastic. Mr. Grundherr has a tremendous depth of knowledge as to the training, and then the deployment, of disaster-oriented SAR-dogs. Both Mr. Grundherr and Kathy Jo gave 150% of their time and effort in presenting this seminar and no question was too trivial to answer. Mr. Grundherr always had several suggestions for resolving each training problem presented to him during the seminar. Such knowledge, incorporated as it was in a logical, step-by-step training program, is clearly the product of many years in the field and of working with many different types of dogs. As participants, we ALL were very fortunate to attend this seminar and we all sincerely thank he and Kathy Jo for taking the time to share their expertise with us throughout the seminar.

The members of NE TF 1 and MO TF 1 graciously allowed me to attend this seminar (and extended flawless Midwest-hospitality to me throughout the weekend). To help repay their kindness, the following bullet entries are derived from notes which I took during the seminar. I have taken the liberty of expanding various comments made by Mr. Grundherr by blending in some of my own training thoughts in an effort to maximize the training-benefits offered by this summary.

Hopefully, this synopsis will serve as a reminder of the many valuable comments which Mr. Grundherr and Kathy Jo each made and will help capture how worthwhile this seminar was for all participants. May it also help create other happy SAR-dogs!!

June 17, 1998

Thank you!

Gail McCarthy

TABLE OF CONTENTS

GENERAL COMMENTS
WORK ETHIC
REWARD RELATIONSHIP
CUES
USE OF FOOD AS A TRAINING AID
RUBBLE PILES
RUBBLE HOLES
BARK BARREL/DOOR
TRAINING PUPPIES
TRAINING PROGRESSION
PRIMARY SCENT VS FRINGE SCENT
PENETRATION/DIGGING
RANGING
HOW MUCH TRAINING IS TOO MUCH (or TOO LITTLE)
DINKING OR OTHERWISE FAILING TO ENGAGE THE VICTIM
SOFT DOGS
HYPER DOGS
REAL MISSIONS
ALWAYS, ALWAYS, ALWAYS ENJOY YOUR DOG AND MAKE TRAINING FUN!

GENERAL COMMENTS

1) There is no one training program that will work for all dogs. Rather, the training program should be tailored to meet the needs of the particular dog with which the handler is working. There are, however, certain training objectives and goals which MUST be met for ALL dogs in order to produce the motivated and useful disaster SAR-dog. These objectives include: (i) focused bark alert indicating live human scent, preferably with penetration ("Bark/Penetration Alert"); (ii) confidence on the rubble; (iii) strong reward relationship; (iv) good work ethic; and (v) task endurance.

2) Handlers should not just look at an Advanced dog and "see the final picture" and assume that their own dog will automatically perform at this same level and in the same manner as that particular Advanced dog. It takes many years of training to produce the Advanced SAR-dog and, from working with many dogs over the years, it is clear that a proper training program consists of a series of slow, gradual steps which, after much time, leads to the finished product. Rushing the dog (especially the talented dog) through the initial steps may be exciting for the handler but it will create many problems later. Step-by-step training and consistency is the key to all good SAR-dogs.

3) Handlers should never take big steps with the dog, rather they should take many little gradual steps and establish a good training foundation. Handlers should set up simple problems and repeat the same problem until the dog works the problem well. Do NOT keep setting up new problems if the dog's work-product on the previous problem is only marginal. Get one problem very good and quit.

4) Do not train just to pass the Basic Test. Train because you believe that the dog can pass the Advanced Test.

5) Dogs use their eyes to recognize people, not by their body shape (ie, static), but by how they move (ie, dynamic). People recognize things more by static shape then by movements.

6) Dogs like a routine and they like a place in which they can relax (and feel safe) after they work. Good handlers capitalize on these traits by "putting the dog up" immediately after it works. A crate works best for this but a private car is ok too. Once the dog makes its find and is thoroughly rewarded, let the dog carry its toy (ie, the reward it has earned) back to its crate to relax. (NOTE: If the dog chews its toys, do not leave the toy with the dog if the dog will be left unattended in its crate).

8) When setting up a new and complicated problem, use the most experienced dog first so that the handler can work out any bugs in the problem BEFORE asking the less experienced dogs to try to solve it. (NOTE: If a handler thinks a problem being set up is too difficult for his/her dog and its training level, the handler should skip it or, even better, break it down into several smaller problems).

9) Good victims are very important as they can tell the handler things that the handler does not see about the dog.

10) It is important to take collars off the dog when working hard, complicated or hazardous problems but keep collars on for simpler problems because it is a waste of time to keep having to catch the dog to hook it back up. Also, the constant fussing with the collar can distract the dog.

12) Enjoy your dog and the training of your dog. Be happy for others who have good dogs, even if that dog happens to be better than your dog. Some dogs take longer to train than others. Be supportive of all your team members and of their dogs.

1) This is the dog's willingness to take direction and to work with the handler. A dog with good work ethic maintains focus on its handler and continues to try something when directed to do so by the handler.

1) The handler MUST get the dog to want -- even LUST -- for its reward. (NOTE: In training, the handler MUST make sure to balance "assertiveness" with GOOD MANNERS so that the dog does not become nippy. SAR-training taps heavily into the natural prey drive of a dog and creates assertive dogs. Therefore, the dog MUST start out with a healthy and strong respect for the sanctity of human skin! As a puppy, teach the dog what "ouch" means!!!!)

3) "Victim Loyalty" is the desire (preferably an extreme desire) on the part of a dog for a toy (or food or a combination of both) which the dog has learned through proper conditioning will come from the victim when it performs the appropriate learned-behavior.

4) The handler MUST really want to reward his or her dog and MUST be able to communicate to the dog a genuine feeling of liking the dog and its work-product. If the handler does not feel this, or does not want to work with the dog, the dog will sense this and not work for that handler.

5) Make sure you praise the dog ENTHUSIASTICALLY!! If the handler or victim does not feel like a fool when he or she is rewarding the dog, then the reward is too little and is of little worth to the dog.

6) The amount of time that the victim spends with someone else's dog should be meaningful. Do not be cheap when rewarding someone else's dog!! Remember, that dog's owner will likely be working with your dog next!

7) Dogs can have many toys at home, but it helps the dog if the handler has one special toy that the dog REALLY likes and reserves that toy for SAR-training only.

8) When rewarding the dog, it is important that there be a close and intense interaction with the victim. The interaction must be of sufficient duration to be meaningful to the dog. The dog MUST learn that it is the victim from whom all good things come and that the victim will always reward the dog in a meaningful way. Therefore, the handler should supply the victim with a toy that the victim can hold so that the dog is kept near him or her while being rewarded (ie, tug toy is best). Tennis balls on a string or woven into tug-toys are ok. Loose tennis balls are not good if the dog runs away after snatching one ball. Also, loose tennis balls often get lost in the rubble pile and then the dog's focus is lost as it worries about where the ball went. Plus, lost tennis balls can distract other SAR-dogs.

9) For dogs that get disinterested in the reward, tease the dog with the toy and then do not let them have it. This teasing makes the dog sometimes want the toy more. Doing this takes sensitivity as you do not want the dog to think that it is never going to get the toy and quit trying.

10) Sometimes the victim should tease the dog with the toy but not let the dog have the toy. The handler should then walk up to dog, snap on the leash and walk the dog away from the victim as the victim continues to tease the dog. The beginning dog can be sent right back in to work this same problem but a more advanced dog can be put away and then asked to work this same problem the next day. This helps create drive and frustration. (NOTE: Preventing the dog from getting the reward often increases the dog's drive for the reward. This also applies to we humans too!)

11) To help keep the dog from thinking that the dog itself can make the decision as to "when it can leave the rubble pile," the handler can calmly go up to dog once the dog has made its find (and is in the process of alerting) and "walk the dog off the pile." The dog can be walked off the pile (i) after the dog has been given the toy by the victim or (ii) as the victim continues to tease the dog but does not actually give the toy to the dog. By walking the dog off the pile as the handler continues to reward it or the victim continues to tease it, the dog learns he only leaves the pile when the handler comes alongside to walk him off.

1) There are MANY cues we can use to help communicate to our dogs what it is that we want them to do. Some of the cues are:

*scent cues - smell of a well-used toy or smell of food, etc (this is similar to AKC ground tracking where the handler uses hotdogs to teach step-by-step tracking in the early stages of training); human scent is a permanent cue to a trained SAR-dog

*auditory cues - calling the dog, praising the dog, tapping the bark barrel, softly coughing etc

*visual cues - seeing the toy, seeing the victim's body (especially eyes and body position (cringing or ducking)), seeing the bark barrel or the bark barrel lid, seeing the rubble pile etc

2) Human scent will remain a permanent "scent cue" for the dog. Similarly, the rubble pile itself will always remain as one big "visual cue" but ALL other cues should eventually be faded as the dog gradually understands what it is that we want it to do and is motivated to do it. (Toys and food can be put in plastic bags and the front of a bark barrel can eventually be camouflaged). A good SAR-dog must INDEPENDENTLY (ie, in the absence of cues or handler support) seek human scent which it has learned to associate with getting its reward.

3) Depending on the dog, some cues are easier to fade than other cues so one should always balance the dogs "need" for the cue against the dog developing a permanent "reliance" on the cue (which should be avoided).

4) One needs to be very careful so that you do not create cue-dependent dogs. For example, some dogs "key off" of seeing the victim with the toy and will not bark without this cue. Therefore, you must be careful to fade such a cue out as soon as you can. This is not to say that you will never use this cue in your training. All good training is a balancing of "what the dog needs" in order to "get the behavior to happen" (ie, jump-start the behavior) or to maintain the behavior. However, if you need the "cue" to initiate the target-behavior ALL the time, then a piece of your foundation is missing (ie, maybe the dog is not motivated, or maybe the dog is not comfortable or is unsure on the rubble and needs to see the victim for confidence). Whatever the reason, the cue should be used initially to explain to the dog what we want it to do but, as training progresses, the cue should not remain the ultimate reason that the dog does the behavior.

5) Sometimes the victim can "allow" the dog to "win" by verbally rewarding the dog or giving it food AS THE DOG IS PERFORMING the alert. In other words, the victim is rewarding "the act of alerting" while the alert is occurring. Allowing the dog to win in this way often makes the alert stronger. Of course, for some dogs, not rewarding the alert every time makes the alert stronger so you need to know your dog!! (NOTE: CASSDA rarely has the victim talk to the dog (but sometimes CASSDA makes an exception to this, e.g. when it is needed to keep the dog at the hole)).

6) As the bark-barrel door is a big visual cue, add flaps or other things to camouflage the bark-barrel door so that the dog does not visually cue on the lid.

7) REMEMBER: The good handler initially uses "cues, bribes or lures" to "get the behavior to happen" during the early stages of training but the GOAL of training is to "get the behavior to happen INDEPENDENTLY" and then to REWARD the proper behavior when it occurs.

1) When working 2-3 victim problems or a series of 1-victim problems, it is easier on the dog (and the victim!!) if the dog is rewarded with food on the initial problems. Reserve the tug-toy for the last problem. It is easier to use food for the first problems as often the dog will "steal" the toy from the victim and then run off the pile. It is also easier to use food to reward the "Bark/Penetration Alert" as the dog is performing it. The food can also be used to tease the dog as it performs the "Bark/Penetration Alert" in order to create frustration so that the intensity of the "Bark/Penetration Alert" increases.

2) When using food as a reward, use big chunks so that none of the food spills on the ground and contaminates the area.

3) For beginner dogs who are learning to cue on human scent, one can softly blow over the food to make it easier for the dog to catch the scent.

4) For more seasoned dogs that are not accustomed to being rewarded with food, introduce the food during a couple of easy bark barrel repetitions so that the dog starts to look for (and to expect) food. During these repetitions, the handler should start to tease the dog and frustrate the dog with the food as this is how it will be used on the pile.

5) To help protect sensitive human fingers and to help get the food to the dog through the opening in the bark barrel lid, thread several pieces of the food onto long wooden chopsticks. Push one piece of food toward the end of the chopstick when rewarding the dog. The sticks can be used to tease the dog by holding the food right out of reach or blown over to generate stronger scent. (NOTE: When using food, make sure that the dog has some familiarity with the sticks so that it "knows" how to get the food off without impaling itself. Make sure that the sticks are very sturdy so they do not break off easily).

6) If the handler is going to use a toy to play with (and preferably it will be the last problem of the set), the handler should not approach the dog until the victim is through teasing the dog with the toy and has actually given the toy to the dog. The victim should then hold onto the toy until the handler quickly (but calmly) accesses the pile and takes the toy from the victim. The handler should continue to play with the dog as s/he walks the dog off the pile. (NOTE: For dogs that are too handler-oriented, CASSDA often lets the dog have the toy after pulling it for a while. With this, the dog learns that the handler does not have to be close for the dog to get its reward).

1) Rubble piles should be made out of rubble with sewer pipes sunk in various place all through the pile so that there are plenty of places for the victims to hide.

2) In training the beginning disaster-dog, it is ESSENTIAL that the dog be trained in rubble where the scent can be controlled. Lots of dirt should be added internally as the pile is built so that wind cannot flow randomly through the pile. Uncontrolled wind causes fringe scent to spread through the pile and it confuses the beginner dog and teaches it to alert on fringe scent and not primary scent. (NOTE: Be careful using dirt in rubble piles as the dirt can make the rubble pile slippery when wet so only use dirt inside the pile or along the base).

1) When building rubble holes, very important to design them so that the dog will not get hurt when it alerts at the hole.

2) To help the dog make the transition to the rubble pile, start to put rocks, brick, plastic sheets etc all around the ground in front of the bark-barrel once the dog's alert-behavior gets strong and secure. This way the dog gets used to alerting on all kinds of funky surfaces BEFORE asking the dog to perform on the rubble pile itself.

3) Beginning search problems should be very easy and should be very comfortable for the dog. Over time, work up to uncomfortable footing etc.

4) Always, ALWAYS use rubble holes for the beginner and intermediate dogs which allow the scent to be controlled and that limit the scent to come out in one spot. For training, the handler must control how and where the scent comes out -- a dog trained in this way will be less likely to alert when it hits fringe scent in real missions. On the other hand, if scent is not controlled and is allowed to flow through the rubble, the dog will learn to alert on fringe scent. At the very least, the dog will alert where the victim cannot reward it. (NOTE: For the training of inexperienced dogs, it is important that the victim be buried in such a way that allows the victim to reward the dog!)

1) An Enzler door is one that has a half-moon cut out of it and another piece of wood is shaped so it that covers the half-moon cut out. Runners are added so that the piece of wood can be pulled back with a rope. It is also nice if the bark-barrel door can be designed so that it fits snugly within a circle cut out of a sheet of plywood. Glue a rubber gasket to the underside of the plywood so that the plywood can be fitted snugly over the opening of a rubble pipe. (NOTE: Ideally, the slider should be made out of a plastic material that does not warp over time, especially if it gets wet).

1) A disaster SAR-dog should not be asked to walk, play or search on a full-sized rubble pile until it is 9-10 months old as its joints, muscles and bones are too immature before this age. Asking the young dog to work on rubble before it is mature enough to do so safely could permanently injure the dog. Joints can also be damaged so that they wear out at an earlier age.

2) Puppies less than 9-10 months of age can still be introduced to funky footings (ie, chain link fence on ground, crinkly paper and plastics, etc) and asked to navigate over small concrete blocks, bricks and rocks. The puppy can also be asked to do easy puzzle-solving problems for its toy by hiding the toy inside cardboard boxes and asking the dog to navigate through mazes made of furniture etc in order to find its toy.

3) For puppies that already know how to bark on cue (see other materials for how to train this behavior), the handler should teach the puppy to focus on the handler by having someone hold the puppy's leash or tying the puppy to a post and then teasing the puppy with a toy it really likes. (NOTE: Make sure you walk puppy to end of its leash before leaving puppy so it does not jerk its neck). When the puppy is barking for its toy and is really focused on the handler, the handler should throw the toy to the puppy. Over time, the dog can be asked to focus longer or bark more BUT NEVER just make this game harder and harder or the dog will lose interest and quit trying. The dog should always be rewarded more often for barking once or twice than asked to bark for 30 seconds. (NOTE: Good handlers use a VARIABLE SYSTEM OF REINFORCEMENT keyed to his/her dog!!)

3) Some young dogs have a "soft" period around 12-months and get skittish about strange things. (NOTE: This "soft" period typically occurs when the dog reaches puberty: for females, this is around the time of the first heat cycle; for males, it is a little later and usually before they lift their legs to urinate, if that ever happens). If the dog seems distracted during this time period, do not force the dog to perform if it does not want to as the dog could permanently shut down.

2) Generally, the progression one should follow in getting a dog to perform on the rubble pile is: (i) Bark/Penetration Alert at a bark barrel; (ii) Bark/Penetration Alert at bottom of pile; and (iii) Bark/Penetration Alert at concrete pipe in pile.

3) Some of the steps centering just on the bark-barrel could include: (i) open bark hole; (ii) shut bark hole with half-moon cut out to encourage penetration; (iii) make a chute in front of the bark hole by building sides along the left and right edges of the bark hole; (iv) lay one pallet across the chute leading to the bark door; (v) use two pallets which are open at bottom; (vi) use two pallets split in middle; (vii) have no opening between pallets and have the handler walk up and open the pallets so that the dog can access the chute and the bark door; and (viii) place funky footing all around the bark barrel through which the dog must work.

4) To get non-barking dogs to bark, handler can use food as a cue. Let the dog lick the food from the hand and then tease the dog with the food. Crouch down, cringe with the body. If dog is submissive, do not stare at dog.

5) Once dog is readily barking at handler on cue, get the dog to bark at a helper/victim.

6) Once dog barks strongly at the victim and is demanding its toy by barking, have the victim run a short distance and stop. Release dog. Dog should run down and bark at victim. The distance the dog must cover gradually increases as the victim gradually moves closer and closer to the bark barrel as training progresses.

7) Introducing the dog to the bark-barrel should be phased in. For example, the victim should run to the edge of the bark barrel and get down on his/her knees and get the dog to bark. Victim should next crawl backwards into barrel and lie halfway out and have dog bark. Victim should crawl all the way inside and get dog to bark. For very soft dogs, victim can even encourage dog to come inside the bark barrel and bark. (NOTE: ALL dogs benefit from having the victim crawl far back into the barrel and allowing the dog to come inside and bark for a couple of repetitions before the bark barrel lid is used as this helps to encourage penetration).

8) Each time that a new barrel, or a new location, or some other new variable is introduced, the beginner dog should be walked up and allowed an opportunity to examine the new problem. Victim can even call out to the dog before it is asked to work it.

9) At each step of training, the dog should be gradually conditioned to bark multiple times (ie, the dog must learn to bark continuously and not just one bark).

10) To link the "act of barking" with a reward in the dog's mind, the bark barrel door should always be opened "on a bark" so that the dog believes that it is its "assertiveness" which reaps rewards. (NOTE: This is only true for dogs that are not strong in barking but are strong in digging or penetrating. In other words, reward the dog for the weaker behavior, be it barking or penetrating or digging).

11) If the dog starts to bark as soon as it us brought up to the bark barrel for a runaway, this is a good thing as it is a "sign" that the dog is anticipating the behavior and "has got it." Therefore, do not try to prevent a dog who wants to bark as it waits for its release word.

12) To wean dogs off of relying on the cue of "seeing the victim run into the bark barrel," do one problem where the dog is still "primed" first with a runaway, For the next repetition, delete the "runaway prompt" and have the victim lying in the bark barrel with the door open. Walk the dog by the barrel with the victim lying in it. Dog should be able to see the victim in the barrel. Depending on the drive and the focus of the dog, the victim can call out and tease the dog. Continue walking the dog past the bark barrel about 30 feet (or even pull the dog away from the barrel). Turn around and release the dog.

13) Another variation of the "No Runaway Prompt" is for the handler to throw the dog's toy into the victim who is in the barrel. Alternately, the handler can throw the toy to a helper who will walk the toy to the victim in the barrel. Do not release the dog until this helper has walked back and is standing behind the handler and the dog.

14) Getting the dog to bark with "No Runaway Prompt" is an important step and the dog should stay at this step (with the door first open all the way, then partially closed and then all the way closed) until the dog is able to bark in front of a bark barrel with no runaway cuing for varying lengths of time before being rewarded.

15) For more advanced dogs, the handler can gradually improve and develop the dog's confidence and its willingness to follow the scent back to the primary source by spreading distractions all around the opening of the bark barrel. The distractions can be made so that the front of the barrel is ultimately barricaded entirely. All kinds of things can be used to clutter up the face of the bark barrel. One can build a wooden chute by angling boards from the top edge of the barrel down to the ground. (NOTE: The chute should be designed so that the handler can crawl in after the dog and be alongside the dog as it alerts at the bark-barrel door). Pallets and sheets of plywood can be leaned up against the angled wood to form the chute. Initially, the path to the barrel should be kept clear and easy. Gradually, more and more funky things can be laid around so that the dog has to work through things to get to the victim. One pallet can be braced at the top of the angled wood so that the dog will have to go under it in order to get to the door of the bark barrel. The opening that the dog has to go through can be made smaller and smaller so that the dog has to aggressively crawl through in order to get in and to access the bark barrel door. Ultimately the problem can be designed so that the dog cannot get through and the handler will have to open the pallets to let the dog in. (NOTE: What CASSDA has done during two of its recent trainings is to use a plastic barrel or tube right up to the alert-cover. The dogs were sent in through the tunnel and expected to bark inside the tunnel when it arrived at the alert cover. Each dog that was sent in ultimately performed a nice barking alert, some quicker than others and some of them more intensely that others but mostly in the same fashion as they typically alerted without the barrel).

16) Two great encouraging problems that one works up to gradually is to leave a large opening between two pallets which can gradually be narrowed over time. The dog will need to jump through the opening in order to get to primary scent.

17) The handler should expect the performance level of the dog to decrease every time complications are added to a straight-forward problem. Initially, these type of problems can (and will) back the dog off as these problems involve many psychological elements including the dog's ability to puzzle-solve as well as to move quickly from light into dark. (NOTE: This should be repeated with the dog until all hesitation to drive to the barrel is gone).

18) The real test in TRAINING is not to see if the dog does it perfectly the first time, but rather does the dog do it better the second time! It is ok if a dog hesitates on the first attempt to solve a new problem. The key to drive and work ethic is whether the dog "remembers" or "learns" from the first repetition and does it better the second time. Should a dog falter on a problem, however, do not hesitate to back up. One benchmark in TRAINING as to whether a dog is able to move up to harder problems is the dog's ability to solve a problem with no hesitation the first time the dog is presented with the problem.

19) Swiss dog handlers often say that the performance level of their dogs during real searches is not so good the first half-day but that performance improves as the dog becomes more fluent with the pile. (Of course, the handlers are also becoming more comfortable with the pile so it is hard to decide whether nervous handlers or the new rubble pile causes the initial decline in performance).

20) When introducing rubble searches to an inexperienced dog, do a pop-up at a well-made hole in the rubble as a primer for the dog. Then set up a run-away where they victim runs and hides into this same well-made hole and the dog has to work through a bit of rubble to get to the victim. (NOTE: Using the pile and well-know holes in this manner helps develop drive "up and over" the pile). Gradually increase the distance the dog has to work to get to the victim at this same hole. (In other words, the victim should stay at the same spot but the handler increases the distance between the location of the victim and the point where the dog is released). Introduce short blind problems in the same progressive manner. For more advanced dogs, start dog farther and farther back in the rubble so that the dog has to work harder and longer to get to the well-know hole.

21) When introducing rubble searches to an inexperienced dog, put victim in a scent hole that is close to the edge of the rubble pile. In such cases, the scent will not permeate the whole pile but will spill over onto the ground. This helps reduce the chance of the dog alerting on fringe scent. (The dog's footing is also better/easier at the perimeter of the rubble which most often has an effect on the dog's willingness to alert).

22) When introducing rubble searches to an inexperienced dog, another thing that you can do is have the dog standing by the pop-up hole. Have the victim pop-up from the hole and call the dog. Walk dog away about 25 feet. Victim hides. Release the dog.

23) Pop-ups in rubble are better than runaways mainly because runways in the rubble are too time-consuming when you are working multiple problems.

24) Handler should have a nice area for dog to bark during the beginning stages of training and then gradually make footing more trappy on the same problem. RULE: If you add complications, add them in a place where the dog is already working comfortably.

1) When training the disaster SAR-dog, the human scent MUST be contained so that it is released ONLY from one spot. You do NOT want the dog to alert on fringe scent.

1) To encourage penetration, use a door with a half-moon cut out and rotate the opening. If the dog likes to come in high, have the opening near the top of the bark barrel and gradually swivel it until the opening is at the bottom edge of the bark barrel.

2) Dogs should be encouraged to penetrate by allowing the dog to stick its head through the half-moon opening where it will be rewarded (ie, feed the dog with its head through the opening). The dog should be teased and encouraged to bark in this position to communicate that going in and penetrating is a good thing but that the dog should still bark.

3) Sand can also be placed in front of the bark barrel and, if the problem is designed that the dog is not able to stick its head into the barrel, it will get frustrated and start to dig.

4) Some dogs want to be rewarded high in the air but it is best to get them accustomed to being rewarded from the bottom of the barrel. Even on the Advanced Test, try to reward from the bottom near the victim and not up high.

5) Seasoned dogs that jump back and/or dance around the bark barrel opening can be re-conditioned to stay put and bark and to try to penetrate by using the half-moon door combined with food on chopsticks (which should be used to feed the dog with its head in the tube). This may take 3-9 months of careful re-conditioning. (NOTE: Victim must always wait for the dog to bark while in the tube so that the bark alert is maintained).

6) A nice transition step to help the dog understand that the problem is not over just because it barks is to set up a bark-barrel and close the barrel with your Enzler lid. As described above, build a 5-foot chute leading up to the bark barrel. The end of the chute should be designed to that the dog cannot access the interior of the chute. When the dog alerts at the end of the chute, go up to dog and verbally reward it. Remove the barrier and allow the dog to work its way down the chute and bark at the bark barrel door where the victim can reward dog on some pre-arranged number of barks. Over time, the length of the chute can be increased and more funky footing added that the dog must work through.

1) Some dogs do not range well. To increase a young dog's willingness to range away from the handler, play a retrieve game with the dog as follows: (i) Tie dog to a fence; (ii) Tease dog with toy; (iii) Have dog watch you walk out 15 yards and drop toy; and (iv) Return to dog, ask dog to retrieve and release dog. Gradually increase the distance you walk before dropping toy. Alternate which direction (right/left) you ask the dog to search.

1) Always, ALWAYS, ALWAYS leave the dog wanting more! The handler must be careful NOT to sour the dog or burn it out.

2) Make sure that the dog always, ALWAYS, ALWAYS wins!!!! The dog must NEVER lose. If you find that you have over-faced the dog, re-structure the problem so that it quickly has a win!

3) Always "keep the cash register full!!!" Long, complex problems are draining on the dog. Balance a hard problem with easy ones so that the dog does not start to think that SAR-work just gets "harder and harder." In order to maintain a strong desire to work in the dog, good handlers determine how many "easy" problems need to be offered to the dog compared to the "hard or complex" problems For example, some dogs need to work 5 sets of runaways or pop-ups for every long blind search they work in order to remain "eager" in their training. Make sure that you end each training session with an easy search problem (or even just the alert itself) which the dog knows and executes well. This is also good for the handler, with both ending the training session after something went well.

4) Make sure you balance tough demanding trainings with time for the dog "just to be a dog." Let it relax and run around and just play.

5) Some dogs (usually high drive males) need time to unwind before asking the dog to work and ALL dogs need time just to be dogs and to play.

6) When training on the rubble, handlers should gradually work up to 20-45 second bark alert before the handler gives "handler support" either verbally or by running up to pile. When testing, it is a good test-taking skill to "call the alert" after 1-2 barks, but not when training. Expect more performance when training so that you have more performance to rely on in real life. (In other words, train for a harder situation so that you have more in reserve when going into a test or into a real mission).

7) A typical training weekend with California-Swiss Search Dog Association ("CASSDA") consists of two consecutive days of training broken down into 3-5 training sessions each morning and 3-5 training sessions each afternoon. One needs to build to this level of training over a period of months but the dog should be conditioned to work this hard for the FEMA Advanced Test. (Again, the last problem is usually a simple alert problem in a well-known alert hole or a simple perimeter problem after which the dog is fed his evening meal).

8) Three days for one seminar is as much as most people can stay focused. Dogs also work very well for two days but are not as good on the third. (This could be caused by the handlers getting tired as well, something that the dogs can read very well, much more than we like to give them credit).

1) The goal of good training is to set up problems which reflect the dog's current level of training. If the problem is at the proper level for the dog, the dog should be able to solve the problem in a straight-forward methodical, yet intense, way. Over time, some (but not all!) problems should be designed so that they take longer to solve (ie, to build task endurance) but the dog should still be able to solve the problem in a methodical way with intensity. When making long-lasting search problems, especially at the beginning, allow for an easy alert situation so that the dog has a realistic chance of success.

2) If the dog first engages and then leaves the bark barrel or rubble-hole, have the victim reward the dog the moment that it comes back and sniffs.

3) Despite the best of intentions, sometimes things go south and the dog leaves the bark barrel and does not come back at all. Sometimes the dog gets distracted by other sights or smells. Or the dog may be confused and run around frantically and not seem to connect with the scent or be able to follow it back to the source. Or the dog may just shut down and want to return to the handler. Any number of things can happen during a problem but, if it appears that the dog is having problems solving the problem, give the dog a little time to re-start or to work it out SO LONG AS the dog appears to be working and trying to solve the problem. But, do not let this go on very long. If the dog is not successful for whatever reason, then cue, bribe or lure to get the behavior to happen -OR- abort the problem. If you don't do this and you reward the dog when it happens to bumble into the victim and bark after it has been milling around, smelling the roses, peeing on rocks or racing around for 20 minutes after you ask it to search, then all you have done is taught the dog that no matter how poor its performance it will still get reward. It is almost worse, however, to allow the confused dog that IS trying to go on and on without being able to solve the problem because these dogs will ultimately just quit trying.

4) Therefore, the handler should read what is happening with the dog and, before things get too bad, the handler should either (i) abort the problem; or (ii) the victim (not the handler) should "get the behavior to happen" by cuing, bribing or luring. You can wait a little bit to see if the dog will re-focus or solve the problem, but NOT long. Get the behavior to happen somehow (cue, bribe or lure) or abort. In either case, an easier problem should be immediately offered to the dog which it can win on.

5) If the dog is beginning to lose focus or confidence and you want to communicate to the victim to reward the dog on the very next time the dog shows the smallest interest in performing the behavior (ie, sniffing the pop-up hole, small whine or slight bark etc), one of the observers can shout "ALERT" or "Ansike" (which means "indicate" in German). The victim should understand this to mean "reward the dog the next time it comes near!" (NOTE: Anything that the handler does consistently sends a message to the dog so be careful in shouting out to the victim as the dog will learn to perceive this as signal (ie, a "cue") if done too often).

6) For dogs that continue to search and race around but which are inexperienced, the handler (who is off the pile) can "cue" the dog by walking back along the pile to where the dog is working and then slowly start to walk back up the pile towards the vicinity of the victim. Dogs tend to cue off where the handler is and if the handler moves in one direction then the dog will likely move in that direction also. (NOTE: This is a form of "handler support.") Alternately, one could just abort the problem and call out to victim to come out of the hole and do a pop-up. The dog should be worked several times on these pop-ups with the approach to the hole being made progressively harder each time. The victim can then tease the dog but not give the toy at the next to the last problem. The victim can just tease at the door and the handler can come up and pull the dog off the pile by the leash if the terrain allows. The final problem should be the dog being sent in to this same hole without a pop-up.

7) Dogs that have been imprinted first with a "refind alert" (especially those dogs that have been imprinted improperly) have their own unique set of training problems which are beyond the scope of this training-summary.

1) EVERY dog has some weaknesses. There is NO perfect dog. A trainer cannot tell if a dog is going to make it in a training program after one or two sessions. Still, if the basic drives are not present in the dog, the going will prove to be very tough for the handler (and the dog). Some dogs are just not cut out for SAR-work.

2) A strong, well-developed and properly CHANNELLED prey-drive will propel a SAR-dog into solving complex or challenging search problems. Dogs that are "soft" or which have little prey-drive can be "pattern-trained" after much hard work and will perform well on bark-barrels or rubble piles that they know well and are comfortable with. However, when a "soft" dog is asked to perform on a new pile, it will lose this "home-field advantage" and the dog is often back at ground zero. (NOTE: One cannot "pattern-train" for every contingency. At some point, the dog must GENERALIZE and apply the lessons it has learn on SPECIFIC piles to solving problems on strange piles. A "soft" dog is rarely willing to make this training leap and often fails to perform in search situations which are new or unfamiliar to it).

3) There are good training systems available today to the dedicated SAR-dog handler and it is much easier on the handler (AND the SAR-team) to work with a dog that has the drives and the temperament suitable for the training. One should NEVER force a dog to do something which it is afraid to do.

4) A handler will need to spend alot of time building up a soft dog. Only the handler can make the decision as to "how much time is too much time" but some teams have determined that it makes little sense to work with a dog which, after three separate training sessions, does not make significant improvements in its willingness to engage a victim. With such dogs, the handler should be asked whether they really want to do SAR work (which requires a dog that can assert itself for its toy) or whether they just want to work with that "particular dog." If the former is the case, the handler should be encouraged to get another dog. If the latter applies, then maybe the handler needs to find another activity which is more suited to the dog than SAR training.

5) To help the dog keep "focused," the handler can walk the dog in the direction of the bark barrel and release the dog on the "fly." This sometimes makes its easier on the soft dog than a "standing-start." For really soft dogs or low-drive dogs, the handler can be running up to the bark barrel from far way and then release the dog in motion.

6) If a dog keeps looking back at its handler for support when it is barking at the bark barrel or hole, a helper should walk up and stand in front of the handler for longer and longer intervals so that the dog gets accustomed to not seeing the handler.

7) Problems like "looking back," will not be corrected after one training session but only after many, many correct repetitions so that the correct behavior becomes a habit.

8) Performing sets of 3 problems is a common rule but is not absolute. Rather, the rule should be whether the dog succeeded on the problem and, if so, should training end with that problem or is there something more that can be gained by doing another problem.

1) It sometimes helps dogs which get very hyper or excited during training to be brought to the rubble and then back to the car on a leash. The goal for these type of dogs is "to focus." Allowing these dogs to frantically race around before or after a search can undermine the hyper dog's ability to concentrate during a search. When the problem is finished, reward the dog and then walk it on the leash back to the crate or car. This will help settle the dog and help keep it focused. (NOTE: Not allowing the dog to play intensely OFF the rubble will also prevent the dog from learning that it has more "non-stressful" fun OFF the rubble pile (ie, the dog should learn that being ON the pile is fun in and of itself!!) In most cases, these dogs also work better if they have a chance to run around in an open field before the training session starts -- in other words, let these dogs get some of the internal steam out!)

1) The Swiss do not give out which dog makes the find. The Swiss keep this information within the unit. The Swiss believe that it is the TEAM that enables any and all finds to be made. Therefore, it is the whole TEAM that should be congratulated on the TV news-show and not just the team member that was lucky enough to be assigned the "right" sector at the right time.

2) Add powdered milk of condensed milk to dog's water to get it to drink on real missions.

3) In conditions of "no wind," a dog may not hit on scent that is further than 32 feet away (ie, 10 yards). Therefore, you need to work closer grid lines when the wind is poor.

4) If people realistically look at the sheer magnitude of the destruction around which (and up on which) disaster dogs are expected to work, then the question of "bark alert" versus "refind dog" is a non-starter. Who could EVER want their dog doing a refind and having it run up and back down and then up again on the face of 10-story pancaked building? It is also extremely unlikely that refind dogs will show the penetrating behavior which is so important in disaster dogs!

5) During deployment, the best way to work a pancaked 10-story building is to search the perimeter first and then the top of the rubble pile if the handler and that dog can get on top, especially if weather conditions are such that the scent is rising.

6) In real missions, the dog will often not work up to par (especially during the first few hours) but the handler should be able to read his or her dog so that the handler can tell when the dog has found even if the dog has not alerted in its usual manner. The handler must be able to read the dog and not just wait for the dog to perform its bark-alert as the alert-behavior of the dog may be far below its normal performance level when searching during a real mission. (NOTE: Sometimes, even on a real mission, the handler will have to set up a little, easy search/alert problem for the dog. This helps the dog and its handler to build up their confidence. Naturally, this should be done where only the dog handlers can see it (but this is often hard to achieve with all the chaos of a real emergency response event). Being sensitive to the need for these type of problems, and setting them up for handlers who may be stressed and not as objective as normal in judging their dog's performance, is one of the tasks of a good Search Team Manager. The Swiss have long been aware of the importance of a very dog-savvy Search Team Manager and conduct lengthy and thorough training sessions for their Search Team Managers. In a real missions, Search Team Managers or their Assistant may be an actual dog handler who does not have a mission-ready dog at the time of the response.

7) NO DOG IS A TOP PERFORMER IN EVERY ASPECT. Some dogs are terrific searchers but the alert of many of these dogs is not as intense as it should be. Some dogs are fair searchers but excellent alerters. Very few dogs are good at both. Some dogs are better climbers than others. Some dogs are better suited for free searches. Others are better suited for fine searches. When putting the optimal team of dog/handlers together, one should try to get as good a mix as possible. If known, select the dogs according to the situation that is going to be encountered.

10) A seasoned SAR-dog will remember where the victim was last seen during a runaway or a pop-up but "out-of-sight is out-of-mind" to the inexperienced dog. So, even very easy problems can become major search problems for the inexperienced dog who literally is "searching" for the victim even though the victim may have ducked down just seconds before the dog is released. THEREFORE, make it easy for the beginner dog to solve a problem by having the victim hide in places that the dog can easily find. Everything should be ENCOURAGING for the beginner dog. Nothing should be DISCOURAGING. (NOTE: Remember that if you release the dog immediately after the victim has ducked down, very little scent is available to the dog should it not "remember" where the victim is -- therefore, quick runaways can be VERY difficult for the beginning dog to solve!)

4) It is important that dogs with a strong drive to penetrate do not learn to think that penetration is the preferred behavior. Make sure that your victim ALWAYS rewards on a bark and gets the dog to bark even if the victim has to cue the dog to get it to understand that barking is the preferred behavior.