This course was about wearables and e-textiles in the field of gardening.

We started our research by visiting a professional gardening school and interviewing the main gardening trainer.

Our first impression at the gardening school was, that gardeners have to interact with many objects in the course of one day. So for a wearable or e-textile to be useful it needs to be useful in more then only one particular situation. Further findings were:

• There are many things that need to be controlled like humidity, lighting, water etc.
• Also we found that forgetting tasks or plants can be an issue.
• Computer are used but seen as not really appropriate interfaces / even distractive

Our first sketches were dealing with the idea to augment the gardeners environment with additional senses for urgent tasks, problems and alike. In the interviews the gardeners said many times that it would be great if the plants could just talk to them and tell them when they need something.

Copyright

©Nikita Jerschov

We started the first Prototype by just testing out different Feedback-possibilities. We conducted a small user test about the tactile feedback patterns to have an idea how distinguishable tactile feedback would be. We found that a small amount of distinguishable feedbacks are well recognized.

After the first tests and the first prototype we continued with the ideation phase and identified two main purposes for a wearable in gardening:

• sensing

and

• in place controlling

The sensing takes into account what gardeners have told us about plants that could communicate their needs. The controlling takes into account the multi-modal environment and work-days of gardeners. Also controlling on the spot - without the need to find the switch - would be desirable since switches and buttons can be very far away in certain situations (in big greenhouses for example) or very oddly placed (water-tables had the controls near the floor).

Copyright

©Nikita Jerschov

This prototype was focussing on the „in place controlling“.

We developed a gesture that a user would hardly perform unintended. With that gesture the user is able to „log-in“ to a „controllable context“ (in the picture the device that gets controlled is in the background).

Before he is able to „log-in“ he gets notified by vibration-feedback that he has the option to login. He also gets constant feedback for his input and also for „logging out“. He can perform the same gesture again to „logout“.

Copyright

©Nikita Jerschov

As already found out in our initial research a wearable in gardening would need to adapt to more then one situation. Therefore we developed a „context“-definition consisting of 3 different „contexts“.

immediate interaction: Intended for secure and clear yes/no -questions. It is applicable for situations such as opening doors or verifying identification. Brief vibration is given to initiate the query, which can be followed by bend of the wrist for answer.

Continuous interaction: Intended for continuous interaction with analog control. Is applicable for situations such as controlling the waterflow in a water hose or adjusting blinds. Access is similar to immediate interaction, and control is mapped to the wrist rotation. Action is ended with wrist bending, similar to access.

Sensing: Intended for ubiquitous information. The garden is flooded with data, which is translated to different patterns of vibrotactile feedback depending on the location and the local situation. No user-action is needed.

Copyright

©Nikita Jerschov

All contexts we controlled via radio, although the „handshake“ utilized a infrared-transmitted code for proximity sensing.

(The very small device is a pure sensing-device, that could be placed on plants for sensing their condition)

Copyright

©Nikita Jerschov

The second prototype was build to test our ideas. Therefore we tried to hide most of the technique inside of a textile band because we found that the first prototype was drawing too much attention from the users towards the technique.

User Tests

Our findings were:

good:

• gesture was easy to understand
• feedback and sensation were convincing
• people had a lot of ideas where they would want to use a device like that

bad:

• difficult to put on
• gesture recognition only worked in a certain position due to the limitations of only one accelerometer in the prototype
• prototype felt still to fragile

Copyright

©Nikita Jerschov

One of our findings was that we need to improve the prototype because the limitations of the prototype were limiting the results of the user tests. Especially the gesture recognition needed improvement. To achieve this we developed a specifically for this task suited textile-sensor and added a second accelerometer to the mix.

We then designed a final draft for implementing this in our final prototype.

So what we improved in the last prototype:

• more robust
• gesture recognition worked now in nearly all positions and utilizes additionally a textile stretch sensor that also provided better feedback
• better general feedback

Like this the prototype was much more convincing. When testing the gesture recognition was recognizing the gesture in nearly all positions.

The prototype also felt way more robust and people could act more freely when trying it out.

The prototype was presented in a presentation to the researchers of the „Design Research Lab Berlin“ @ UDK Berlin.

PDF Copyright

©Nikita Jerschov

We went back to the gardening school and interviewed the main gardening teacher again and showed her our prototype together with the prototype-contexts.

Here some observations / findings:

Observed User Experience with our Prototype

• Basic concept was fast understood - login and controlling worked well.

• controlling by rotating the wrist was coherent and directly accepted by the test-person.

• control-range has been a big fascination-factor! The advantages of having a control-range was pointed out by the test-person as a huge advantage.

• login-gesture was experienced as „hard to do“. After explaining „why“ (no unintended logins) it was immediately well accepted and even appreciated.

• The test-person was able to understand the general system and act on it - login/out, control and sense.

Aside from our observation the system was well appreciated by the interview-partner at the gardening school and many links and ideas were developed and documented. We found that there are many applications for a (productivity) wearable in gardening, what we showed with out prototype.

It is self-explanatory that a lot more research and testing would be needed to build/develop a system like that. We could however find interesting applications and also develop some ideas for a context-aware wearable.

We also documented our work in a scientific poster-paper that was submitted to CHI 2015.