Sniff

Combining motherhood and publicity.. The exhibition at DogA led to a feature in the inflight magazine of the Norwegian air craft company ‘Norwegian’. In the middle of the photo shoot my little one woke up and wanted to be with her mummy. It surely gives the article a special twist : )

Dagens Næringsliv publishes an article about the winners of the Young Talent awards. The photo is captured in my appartment with an old Hasselblad-lookalike (I think it was a Pentax), a not so commonly seen equipment among photographers these days. Nice photo.

Mina was born on the 18th of march, wich has lead to a pause in the development process. The following posts documents some of the publicity that naturally follows the award.

Sniff has won a prize for Design for All at Unge Talenter 2008. The newly designed sounds are tried out at the following exhibition. Sadly, the hardware with the mp3 player does not work properly. The prototype have to have a steady power supply, and cannot run on batteries without getting overheated. This is of course a hassle when exhibiting, but it can be solved with a cord running through a dog lead to a power station that is hidden under the display base at the exhibition stand. The lead is good to have anyway, for security reasons to prevent visitors from taking Sniff with them.

…

…

…

The issue with the volume of the sounds, that we already experienced with earlier exhibitions, becomes very clear in the setting at DogA. In this big exhibition hall with lots of people present the sounds drown almost completely. Visitors has to put their ear very close to Sniff to hear the feedback. The current hardware is supposed to being used with headphones, and it is not optimal for being connected to a speaker. Even the amplifier that Einar builds does not improve the volume enough. The volume level in exhibitions is particularly tricky since the number of visitors makes the circumstances change throughout a day.

The mp3 player opens up for so many new options. When Sniff could actually react in any way we want, with any possible sound, we’re forced to consider different type of sound feedback more carefully. We do some quick sketches of dog sounds, with free samples on the web that people have recorded from their own dogs, as a starting point. Theese sounds do not have the needed width in meaning and substance, even if the range stretches from soft growling to loud roars it’s mainly barking anyway. This is where the idea of halfway abstracted sounds appear. We could try to produce the sounds our selves, but the idea of getting a puppeteer to act the sounds is far more tempting. Through a Oslo Nye Dukketeater and Unima Norge i find a puppeteer actress, Camilla Tostrup. She likes the project straight away, and has the same thoughts around personality and level of animality as I do. I don’t think there’s a point to try to make realistic dog sounds, rather work with Sniff as a character, to find his own unique voice. This work is still in progress, but some initial recordings has already been in use at an exhibition.

One thing that was certain when developing new prototypes was that the volume needed to be much louder, and that it needed to be adjustable for use in different situations. The volume suitable for use by a few persons in a small room was way different than the needs for volume in an exhibition setting, or with a lot of kids in a kindergarten.

The USB stick with the music files is now placed where it is accessible, right next to the battery opening. This way new sounds can be added on the stick. But the microcontroller still needs to be updated with new code, so it needs to be connected with an external USB port for the Arduino Mini (not included in this prototype). The stopper in soft foam is not an optimal solution, but will do for this purpose.

New shapes for vacuum moulding is produced out of the old piece of cernit clay, and this time we add a flexible neck. This makes the prototype more vulnerable but we agree on that a flexible neck is a must

The inner sandwiches are connected with one of the plastic layers as “neck”. This neck piece is bent to fit the angle of the dogs head. But in the end this solution might be too rigid as well, and another material is then applied. With thick rubber reinforcement to connect the two body parts, the cords still cannot be stretched or twisted in any way that makes it break. The rubber is fastended within the shell parts, it’s actually screwed tight to the inner sandwich. This makes the cords a lot more protected than in the original Sniff.

The stiff all-in-one shell seemed like a good idea when it comes to durability, but it turned out to be devastating for the experience of the prototype. While the electronics are safe inside this capsule, the overall stiffness of the model surely removed almost all lifelike qualities in the dog. The fact that he no longer was able to bend his head, just made him seem plain dead. At this stage it was clear that a lot of the expression in Sniffs face actually comes from the flexible head, that could be positione in a suitable angle. I guess we just had to try it to understand it didn’t work, at all..

Another design issue was the planned one-piece in fabric that were supposed to provide a simpler form of wrapping. It turned out that the placement of the opening for changing batteries (and occationally uploading new code), was very ugly at the bum of the dog. It wasn’t a must for the planned one-piece plastic shield to be wrapped in one piece of fabric, but the separated head-and-body cover suggests a different solution for the battery opening as well. This original solution turned out to be much neater in reality.

The final model is carefully split in the middle. The placement of the cibatool and wood boxes that represent the technology are now visible again. Each side of the model is slightly built up from underneath with a layer of capa-board, to make the vacuum moulded plastic cover enough of the shape to make a clean cut where the two shell-parts meet.

After moving around some small details in the construction we land on a minimum-size for the sandwiches in the head and the body. Theese are then constructed in cibatool, with a small extra margin on 1 mm on each side.

The cibatool blocks are then embedded in cernit clay, a horribly troublesome material to mould and shape, hard and grainy as it is before it gets warmed by a time consuming hand kneading. The good thing about cernit is that it gets hard enough to vacuum mould after it’s been baked in the oven, a quality that dosen’t apply to the other clays avaliable. This is important since the plan is to construct a mould that can be used in a mass production of five prototypes without loosing it’s shape. The hardened cernit can also be polished and sanded to desired measurements afterwards, tand this is where the extra millimeter is important, as I’m shaping the hard cernit to the smallest possible until the corners of the cibatool blocks appears.

The first version of the shape turns out to have too short head to fit the proportions of the body. When wrapped in tempur foam this shape is likely to turn out almost circular, leaving too little space for Sniffs eyes, ears and face. Even though we want Sniff to be as small as possible, this can not be on the behalf of the expression and the proportions, and it’s important that the head is a bit oversized to express a sniffing quality. I therefor add an extra piece of cibatool in the length of the head.