Practice search for Abby

Kelsy performed perfectly in today's practice search.  This surprised me because the wind blew at 28 miles an hour with gusts reported up to 50 MPH.  I started out on the search expecting we would just have a nice walk in the freezing sideways rain and not follow the trail at all.  Not only did she follow the trail, but she stayed on it exactly, not jumping to the opposite side of the street like she has some times in the past.

We worked a trail 22 hours old, at 12:30 on Sunday afternoon.  The trail had been laid with the target dog, Abby, plus two extra dogs and an extra person, I think, for sixteen feet all together.  So, Kelsy was actually following a pack, not just one dog and a handler.  The trail was 2.43 miles long, with 10 turns and at least 25 intersections.  It rained about .25 inches between laying the trail and starting the search.  It was not windy when the trail was laid.  At the time of the search, winds were steady at 28 MPH, with rain for most of the 45 minutes, temps around 41 degrees, 80% humidity.  A portion of the trail, up California Avenue, was entirely concrete for four blocks, with no vegetation whatsoever, and the street ran north/south in line with the scouring, gale-force winds.  That section of California Ave had high pedestrian traffic and many dogs being walked.  Kelsy had no trouble at all in this section.

About a month ago, when looking for the white terrier in White Center, I had a thought about dead end branches of the trail.  Kelsy seemed to take them all, which takes up a lot of time.  We could gain ground on our quarry if she would skip the dead end branches that go out and double back.  In thinking about why she might take them, it occurred to me that the dead end branch has twice the scent, making it seem like the fresher scent trail, maybe.  The trail we worked today had two double trails, one a dead end and one the combined start/finish.  At the dead end branch, Kelsy took the dead end like I thought she would, and we followed it all the way to the end indicated on the map.  At the end, I thought she would take the double trail down the alley to the end where Abby was waiting.  Instead, she got started on the beginning of the trail again.  Because people were waiting for me in the rain, I redirected Kelsy down the alley, even though she wasn't wrong.  I will keep collecting more data on dead end branches.

Kelsy followed this trail perfectly.  At about the halfway point, she came upon a golden retriever sleeping on the sidewalk in front of a coffee shop.  She didn't speed up when she saw this dog, and she only gave this decoy a quick sniff on the way by, mostly ignoring her, like she should.  I did not have to give Kelsy any corrections other than that I stopped her at intersections while waiting for traffic.  I would be really excited about how well she did, except that having several dogs on the trail might have made it unfairly easy for her.

One very valuable bit of information is that we now know that a dog can follow a scent trail 22 hours old in winds clocked at 28 miles an hour.  It was not windy when the trail was laid, which seems significant to me.

On this training exercise, I made extensive use of my smart phone.  I could have done everything without it, but it made many things easier.

1. I recorded my mileage with an app.
2. I recorded my volunteer hours with the same app, and I can email a report to whomever.
3. I used the map function to find the address of the starting point.
4. I used the phone as an actual phone!
5. Took a picture of the map Jenny had drawn, and the picture was easy to read long after the paper map had become a soggy mess. 
6. Used a different mapping program to record our track, along with total mileage and speed.
7. Took video of Kelsy "in the groove" using her nose and not her eyes, and another video of her ignoring the decoy dog.
8. Recorded voice notes for details of the search and conditions.
9. Received up to the minute weather data to record for the search.
10. Used the map search to find a restaurant for lunch.
11. Used the map search to find the back entrance to Westcrest off-leash park for Kelsy's reward play later.  
12. Checked craigslist to see if I should be on the look out for any missing dogs in the area.
13. Checked email to see if any other volunteers had signed up for target dog duty. 

All of those functions make training easier, but I can use all of those functions on actual searches as well.  True, I got my smart phone as a toy, somewhat, but it really does make life easier some times.

All in all, it was a perfect day, about as much fun as you can have in sideways rain at 41 degrees.  Kelsy was rewarded with cheese, with a little bit of fetch at the end of the trail, with a bite of my sandwich at the restaurant, and with plenty of fetching at the off-leash park.  Plus she found her dog, which is rewarding to her.  And I didn't have to correct her at all, or sigh those sighs like I do sometimes.  A perfect day.

Added note:  Two days later I had an idea about one reason this search may have gone so well:  I didn't talk.  I can think back to several exercises and searches where I had someone, a volunteer or a client, along with me, and I did a lot of talking about what Kelsy was doing and why.  On a practice search for Glenna in Federal Way, I know I didn't talk at all, and that search went perfectly.  I don't know if this is just a coincidence, but if it is significant, I can think of a couple of reasons.  The first reason is that Kelsy does listen to what I say.  That doesn't mean she always obeys, but she listens.  If I start talking on the phone, she usually comes around and wants to do something. I know that I have to avoid saying certain words if I don't want her to start barking in anticipation of some activity.  It might be that I distract her when I talk.  Another reason that talking might distract her is that it means there is someone with us, quite often a person with the scent of the dog we are searching for.  She might think, that person smells like the dog, so she must know where the dog is, so I don't need to work too hard. 

If it is true that talking degrades her search performance, I can either not talk on searches, or talk more in training so that she learns to ignore me.  I'm thinking that I will lean toward not talking on searches.  I will have the owner stay back ten feet and hold questions until later.  Or text me.  I will do more testing on this theory, and see if Kelsy scores higher on silent searches. 

Search for Grover

Pictured above are Grover, Jill, and Chris, moments after Grover came back to the point of escape.  Grover escaped out the front door last night, and his family and friends spent all night looking for him.  They put up flyers around the neighborhood, and they received five sightings.  The most recent sighting was this afternoon around 1, just before they called MPP.  I told them they would most likely find their dog without the help of MPP, but that Kelsy and I would be willing to do a search if they wanted.  They did want all the help they could get, and they wanted to do everything they could to find Grover fast.  Kelsy and I started our search at about 3 PM, at Alder St. and Lake Washington Blvd in the Leschi neighborhood of Seattle.  The most recent sighting, two hours earlier, was of Grover heading down Alder toward Lake Washington Blvd.  Kelsy took the scent from a towel that Grover always rode on in their car.  She started down the hill, but then turned around and followed up the hill.

Kelsy tracked for about 2.5 miles over the next hour.  The conditions were perfect, about 50 degrees, almost no wind, and the ground was damp from recent rains.  We stopped tracking when we ran into private property, about 1.5 miles south of the point of escape.  About an hour later, Jill got a call of a sighting from two hours earlier, on Lake Washington Blvd.  We could not pinpoint the exact location based on the person's description, and when I tried Kelsy on the scent at the point we thought he was talking about, she gave her negative signal.  After that, we received a new call of recent sighting five houses away from the point of escape.  I started Kelsy on the scent trail there, and she indicated that Grover was everywhere.  Kelsy went up and down a long flight of stairs, through people's yards and all around the block, but she could not pinpoint Grover.  I told Jill and Chris that Grover was most likely nearby, but that the looping and overlapping trails made it difficult for Kelsy to pinpoint his location.  I advised them to hang out on the front porch and talk to each other in a normal conversational tone, and Grover would likely come home to them.

Kelsy and I went to the truck, and I waited there for about five minutes while I checked my email and made some notes on the case.  Just as I was about to leave, their daughter knocked on the window of my truck and said that Grover had come home!  I got out and saw Grover a few houses away, on the sidewalk.  He was cautious, looking at them but not advancing.  Jill and Chris lay flat on the ground, following the advice I had given them, and instead of shouting Grover's name, they just talked softly.  I got to see Grover come up to Chris cautiously, and then throw himself onto Chris once he got the positive scent of his owner.  Grover was very happy to be home, and he seemed to be in excellent health.

Jill and Chris graciously agreed to let Kelsy find Grover in a short practice run.  Jill hid with Grover about five houses away, and Kelsy took the scent and charged right up to Grover.  I gave Kelsy lots of treats and praise.  This is how I would like to end every search.  Of course, I would like Kelsy to find the dog every time, but short of that, I would like her to find a dog, even a planted dog, to keep her morale up and reinforce her motivation.

Looking back on the search path, I would say it is possible Kelsy was right all the way, but there is also a good chance that she had jumped from Grover's scent to the scent of another dog for about the last half mile.  I can't say for sure that Kelsy was on the right scent the whole time.

I still think Jill and Chris would likely have found Grover without the help of Kelsy and MPP, but I do think that our advice, the confirmation of the search dog, and our presence helped.  Most of all, I was happy to be there at the  moment when Grover was reunited with his family.

Training with Harley

Harley is the World's Greatest Cat Detection Dog, besides being an all around wonderful dog.

Yesterday, Kelsy found him in a training exercise.  The trail was about four hours old, and about one mile long.  The weather was cool and cloudy, about 47 degrees, with light winds, no rain, and damp ground from recent rains.  Kelsy used a scent article that had been stored in the freezer for 6 months.  For the first quarter mile, which was mostly paved, Kelsy had trouble staying on the scent trail.  She was frequently distracted, and she needed to be guided onto the correct trail.  Once we got through the apartment complex and onto the trail through the woods, Kelsy stayed on the correct trail most of the rest of the way.  At one dead end, she went down the short street where Harley had not gone, and then came back to the trail.  At a point about two blocks from Harley's house, after we stopped and allowed Amy to call ahead, Kelsy sprinted the rest of the way, as fast as she could drag me along.  During this training exercise, I made several stops for Kelsy to rest her nose, so she could smell the scent better.  She is still not staying on the right trail 100%.  I am having trouble coordinating with volunteers to arrange training sessions, but I will have to increase her training frequency in order to get over this problem of following the wrong trail. 

The search for Dexter

Someone broke into Dexter's house last night, robbing the home, and leaving the door and the gate open.  Kelsy started her search about 20 hours after the probable time of escape.  She used a pillow that Dexter slept on as her scent article.  Kelsy spent a long time, probably about fifteen seconds, analyzing the pillow, sniffing all over the surface.  Then she started on the trail pretty decisively.  At the end of the driveway, she turned north for about twenty feet, ran out of scent, and turned south along 108th Ave SE.  Passing the yard of Dexter's house, she was very interested in the fence line, which makes sense.  She made her way south at a steady pace for four blocks, using her nose and not her eyes.  (When Kelsy first started training for search work, she ran everywhere at a gallop, dragging me along, but now she gets into a groove where she is walking/trotting, and the tips of her ears have a signature bob, while her head is held just slightly lower than level.)  Then she followed the scent trail west through a parking lot and onto SE 216th Street.  She started pulling harder after a block, indicating to me that Dexter had rested for a while and then started off again later, leaving a fresher scent trail.  Kelsy crossed the street and then followed the scent trail in the oncoming lane of traffic, which is not the place someone would be walking a dog.  I had to make her wait several times while waiting for traffic to clear, and then she continued on the same trail.  Near 103rd SE, she turned toward an area in the tall grass where an animal had nested for a while, and she sniffed around some shards of smashed pumpkin.  Kelsy followed the scent trail to the end of 103rd, to a house, and did not find any trail leading away from the house.  No one was home.  Kelsy was very interested in a spot in the back yard, where the grass looked like there might have been a scuffle.  Dexter used to get into fights, but hadn't recently.  The people obviously owned more than one dog, but none of the dogs were home at the time.  We left a flier on their door.  I will be very interested to learn if Kelsy was correct, possibly correct, or definitely wrong in the trail she followed.  My assessment of her behavior leads me to judge that I am about 85% sure she was on the right trail. 

The conditions were perfect, overcast, light rain, little wind, about 48 degrees, and the landscape was very damp from recent rains.  Although the area was near a busy four lane highway, the trail Kelsy followed was bordered by grass and brush all the way.  Kelsy started her trail about 3 PM, and the trail ran about half a mile. 

Based on something I had read about dogs becoming desensitized to a smell after too much exposure, I had decided to rest Kelsy frequently on future searches in order to refresh her nose.  I had forgotten about that on this search, but I ended up doing it anyway because we had to wait for traffic several times. 

Butters found, but not by Kelsy

The good news is that Butters has been found and returned safely.  The not-so-good news is that Kelsy was following the wrong dog, again, and I didn't know it.  If Kelsy and I are to be of much use, we need to work through this problem. 

Admittedly, Kelsy didn't have great material to work with.  For a scent article, she used the back seat of a car where Butters would always ride.  Kelsy got the scent of a dog, and she followed the trail of a white dog, according to the reported sighting.  Butters had been picked up Sunday night, so the dog she followed couldn't have been Butters.  Somehow, I need to communicate to Kelsy that I need her to follow the scent of just this dog, not the next best thing. 

What got Butters home was our large signs.  The person who took Butters home saw a large sign a few blocks away, and called the owner. 

Search for Butters

Kelsy searched for Butters today.  He is a West Highland Terrier, about 2.5 years old, all white, about 20 pounds.  He got out an open gate two days ago, at 17th SW and SW 98th Street.  The owner got a sighting today at 11 AM at SW 106th and 17th SW.  Kelsy started on the trail at about 2:15 PM.  Unlike our attempt yesterday, when the winds were as high as 50 MPH, Kelsy seemed to have a good scent.  She followed the trail down the alley to 112th, then west to where 112th dead-ends just past 29th Ave SW.  She came back to 29th and followed south to where 29th dead-ends at Salmon Creek ravine.  Then Kelsy came back up to 116th and headed east to the elementary school, where the trail ended.  The trail ended at a locked gate by the playground at Shorewood Elementary, at SW 116th and 27th Ave SW.  Kelsy tracked for about 1.5 miles.  For most of the time, she held her head in the "I'm in the zone" position, slightly below level, with the tips of her ears bobbing up and out with each step, and not looking around.  At the three dead-end turnarounds, she did give me the negative sign, turning and looking up at me.

Along the route, Kelsy found many interesting things that would have interested Butters as well.  She found spoiled food dumped on the ground, several chicken coops, some small dogs allowed to roam free near their yards, and corn scattered in an alley to feed the crows.  When she found the small dogs, she did not show much interest in them, and she was more interested in following the trail of Butters.  I would say I'm about 90% sure Kelsy was on the right trail.

The winds had calmed to about 5 to 10 MPH, the ground was damp, and it wasn't raining.

Today's trail got me thinking about why Kelsy takes the dead-end trails instead of following the route where the target dog continued on.  It would certainly make the searches faster if she didn't dive into every yard where the missing dog went in and out.  I realized that the scent would be stronger on the dead-end trail instead of the continuing-on trail because the dog walked there twice, once in and once out.  So, when Kelsy comes to an intersection where the dog turned right into a yard and then came back out and continued in the previous direction of travel, the strongest scent is turning right into the dead end, where the dog is not.  I will have to design some practice trails with dead ends to confirm that Kelsy is doing this.  

Decoys

Kelsy trained at Renton Village today, starting at the PIMA school.  Valerie and Pearl had laid a scent trail about 24 hours earlier.  Darlene and Buddy set up as a decoy at the Discount Tire store.  Bill and Bear made a short decoy trail at the Evergreen building.  Sky and her owner stayed at the end of the trail as a final decoy when Kelsy reached the end.  We stated at about 3 PM, with the temp around 48 degrees, little wind, and no rain at the moment.  The ground was damp from recent rains.

Valerie walked with us to tell me when Kelsy went off trail.  For the first 200 yards, Kelsy was exactly on the trail.  When she hit the scent trail of Bear, she accelerated and tried to run to him.  Had this been an actual case, I would have interpreted that acceleration as either a fresh scent of the target dog or excitement over a new scent trail.  Following it, I would have discovered that it either was the target dog or not, and if not, I could go back to the point of divergence.   When we got up to the power line towers and over by Wells Fargo, Kelsy briefly jumped from the out trail to the back trail, and then she crossed over again to the out trail.  The two trails were less than twenty feet apart at that point.  She followed around the Holiday Inn, and I could tell she caught the scent of a little white dog I had seen people walking.  I could tell because her gait changed and she started looking instead of smelling her way along.  When we came to Buddy's decoy trail, she glanced in Buddy's direction, but she didn't veer from Pearl's trail.  Going by a restaurant, she got off the trail by about twenty feet.  Then she zoomed right up to the target area.  She did run up to Sky first, but realized it was not the right dog.  When she charged up to Pearl, who is a puppy, this made Pearl nervous.  After I rewarded Kelsy with treats, she and Pearl got to meet properly, and Pearl relaxed.  Kelsy's treats this time were McNuggets, and she really liked them.  Kelsy stayed on trail over 90% of the time, and the two times she strayed, I could tell that something was different.  Again, I watched for when she was using her nose versus when she was using her eyes.  The trail was about a mile long.  I forgot to set my GPS at the start, so I don't know the exact length.

After Kelsy, I worked Zeke.  I made the mistake of placing the scent article on the dirt, and he was more interested in smelling the dirt than the scent article.  In the future, I should be certain to place the scent article on pavement, to avoid any contamination.  I picked up the scent article and had him smell it in my hand, which seemed to get him started on the right trail.  Zeke works with his nose to the ground much more than Kelsy does.  Zeke stayed right on track most of the time.  He was distracted by the scent of Bear, and I had to pull him back on course.  He cut off a big loop when he jumped across from the outbound to the inbound trail.  When we got close to the end, he saw Karma, who was being a decoy, and he would not go any further until he could check her out.  Once he smelled that she was not the target dog, he continued on.  At the end, he checked out the first dog he saw.  Pearl was hidden, and instead of going to the next visible dog, he followed the scent trail to Pearl's hiding spot.  Zeke did very well.  Reading him is different in many ways than reading Kelsy.

Tess and Kelsy

Training in the rain, again.

On November 6th, Kelsy and I trained with Kat and Kody and Karma and Ryan.  This trail was two days old, near Military Road and S. 268th, at Greg Nelson Park.  The temperature was about 50 degrees, raining off and on, with little wind.  We started trailing Kody at 262nd and 42nd.  Kelsy started out at a walk, and walked most of the time, occasionally breaking into a trot.  Had this been a fresh trail, she would have run most of the time.  On this day, I watched her head, noting if she was using her nose or her eyes.  Most of the time, she was in a groove, just using her nose and plodding along.  When she started to look around, using her eyes instead of her nose, I put tension on the leash.  This kept her away from distractions most of the time.  The trail was about a mile long, and at about the half mile point, Kelsy perked up and took a trail into the woods.  I stopped her after about twenty feet, since this was not the trail Kat had drawn for me on the map.  We got back on track and didn't have any other deviations.  At one intersection, she took some time figuring out the direction of travel, but she used her nose and worked it out.  100 yards from the end, she saw Kat's bright green vest and started running toward it, cheating.  But Kody wasn't with Kat, so Kelsy had to go back to using her nose.  She pulled much harder at this point, being in the fresh scent.  Finally, she saw Ryan's green vest from about twenty feet away, and sprinted to Kody. 

I rewarded her with treats and a game of fetch with a stick.  Had this been an actual search, I feel that Kelsy and I would have done very well.  She went off track once, but I noticed her changed behavior and made note of the point of deviation, so we could get back to business. 

Training Kelsy, October 30th, 2010

Kelsy and I met Kat and Zeke in Federal Way to train on a trail two days old.  At 3 PM it was raining steadily, overcast, with little or no wind.  Kat had walked Kody on the trail two days earlier, and she told me the route ahead of time so that Kelsy and I could try to work out our problems.  Although Kelsy has done well on many cases, on one or two, she has gone off the rails without my being able to tell until much later.  I want to be transparent with anyone who hires us and let them know exactly how we are doing, to the best of my knowledge.

To try to fix our communication problem, we started off with a negative scent trail.  I presented the scent article in a place where Kody had not been, and asked Kelsy to search.  She took off a bit more excitedly than I would have hoped, given the lack of scent, but she gave me the negative after about 100 feet.  Then she checked in several other directions, and gave me her negative signal much quicker.  Her negative signal is to slow down, start walking sideways, and then turn back to look at me.

Next, we started on the scent trail.  I pointed her in the right direction because I didn't want her to just run right back to Kody, standing half a block behind us.  The route was about 1.25 miles long, and Kelsy did well for the first half a mile.  Then she took me into a yard where I know Kat and Kody had not been.  If this had been a search, I might not have known that Kelsy was leading me astray.  Hopefully, it would have been a dead end, and we would have gone back to the right trail.  She increased her pull as she started on the new trail, and I might have interpreted this as getting close to our target.  She did this one more time, speeding up to follow an animal other than Kody.  I stopped her from following the wrong trail, and I would hope this would encourage her to stick to the right trail.  I hope it does not teach her that I know where the lost dog is, and she can rely on me to correct her when she goes astray.

Another thing I noticed while we were searching is that she tends to cast about with her eyes when she is losing the trail.  When she is on track, her head is forward, a bit lower than level, and she is just in a groove.  Another way I have seen her get off track is when her nose works right on the ground in tight little turns.  This usually means a critter.

When we approached Kody's hiding spot, Kelsy burst into overdrive when she entered the fresh smell, as opposed to the two day old scent.  She charged into the park, and took many wrong turns, frantically trying to sort out where Kody was.  I think she was following the footsteps of Kat, who had been in the park working with Karma.  (So many K names.)  I took her back out to the entrance of the park, where she first caught Kody's new scent, and started her again.  This time she got on the right track, but she cheated a little when she saw Dianna in the bushes, holding Kody.  But, she did find Kody, and I rewarded her with treats and fetching, so hopefully this will reinforce the things she did right today.  She was right, 95% of the time.  Those two wrong turns might have made the difference between a successful or failed search.

After Kelsy's turn, I ran Zeke on the same trail, on the scent of Kody.   Zeke did fairly well, although I did have to stop him from taking a couple of wrong turns.  When we got back to the park, he zeroed in on Kody much faster than Kelsy had.  My left knee and right ankle hurt after being dragged for 2.5 miles by two dogs.

Next, Kelsy and Porter and Tess all went to the Grandview off-leash park for a frolic.  Fetching at Grandview is just about Kelsy's favorite thing in the whole world, so I want her to know that a day of searching doesn't mean that she can't go to Grandview.

I wish I had brought a towel.  Note to self....