Filtering our user-centered research experiences into a more refined set of findings and takeaways, we used what we learned to inform us as we brainstormed several installation ideas to present to the exhibition coordinators at the Children’s Museum. Based on their feedback, we derived a robust vision which we could begin prototyping in earnest using an iterative design method. This method is a cyclic process. At each step, the design of a product is refined and prototyped. This prototype is tested, with the evaluation of the results contributing to the refinements in the next step of the cycle. If you’d like to see the final prototype resulting from this process, please visit our homepage.


Each of our visions tried to engage the theme of persistence in The Very Clumsy Click Beetle in a different manner. For example, some focused on the giving and receiving of positive encouragement. Another approach involved setting and achieving of challenging goals. Other visions were about enabling physical play and movement. The ideas that we felt were best at representing these topics were turned into storyboards. The Children’s Museum provided feedback on these, which helped us to develop a final proposal for the installation. Below you can find four of the storyboards that we presented.

Enabling Physical Play and Movement

This installation idea is divided into two components: a screen onto which an animated click beetle is being projected, and a springy ramp. The beetle, stuck on his back, would struggle to right himself but could not do so on his own. By jumping on top of the ramp repeatedly, viewers would be able to launch the click beetle into the air so that it could eventually flip back onto his feet. Feedback: The museum wished to avoid overt use of technology and a reliance on projection screens.

With this proposal we wanted to encourage children to perform a challenging physical task – in this case, a somersault. One half of a wall would show a step-by-step diagram of a click beetle doing a somersault. When children stood in front of the other half, they would see themselves placed in the landscape of the book, the characters encouraging them to somersault. By performing one, children would take the place of the click beetle and better understand the book’s message of persistence. Feedback: An unexpected problem was the fact that many younger children wouldn’t see the diagram as a progression of movements being performed by one click beetle. Instead, they would just see four click beetles.

Setting a Goal

Here, our idea was less centered on the click beetle itself and more focused on a personal challenge. On a wall of the exhibition there would be images of three encouraging animals from the book. For each animal there would be a set of toy click beetles with a small challenge printed on the back – for example, “learn to whistle”. The height the click beetle challenges would be placed at would indicate how complex the task was. This allows taller, older children to reach the more difficult tasks. If children completed the task they chose, they could put it in the “whirler”, a glass apparatus which would blow the click beetles around in celebration. If children could not complete their task, they would place their click beetle back on the wall. The animals would then give them gentle encouragements. Feedback: The idea was more suited to an event or workshop than an installation, as it would require constant supervision on the part of the staff. The “whirler” also seemed a tad anticlimactic as a reward for completing a complex task.

Positive Encouragement

Here we revisited the idea of helping the click beetle flip back onto his feet. A projection on the wall would show several beetles on their backs, all struggling. In the book, the click beetle is visited by a variety of animals that encourage him along the way. Children would take on this role by coming up to the screen and verbally encouraging a beetle of their choice. The more emphatically they encourage, the higher a beetle would flip in the air. If multiple children used the installation, there would be an added competitive element in seeing who could get their beetle to go higher. Feedback: Much like before, the museum wished to avoid overt use of technology and a reliance on projection screens. Furthermore, it wasn’t feasible to program voice recognition capable of consistently differentiating between abusive shouting and encouraging shouting.

Final Idea: Clumsy Clicky Play Area

After receiving feedback, we came up with the idea to create a circular grassy play area which could accommodate 3-5 children at any given time. The area would try to emulate a natural landscape by having slopes and bumps. The whole area would be pressure sensitive. Based on where children stepped, the installation would play a variety of different noises – the rustling of leaves, the whispers of wind, the soft clicks of a click beetle, and words of encouragement suggesting different physical activities for the children to try.

We also thought we could make the children feel more like the clumsy click beetle by playing with their sense of scale. By adding large pillows shaped like natural stones and pool noodles resembling tall reeds of grass, children would feel like the small beetle surrounded by the much larger world around them.


We spent some time building miniature models to get a feel for the shape and arrangement of the play area we would prototype. We used these models to develop a common vision for the exhibit amongst ourselves. They were made using paint, pipe cleaners, and air-drying clay. Underneath the model we added a pressure sensor to simulate the noise-making behavior of the play area.

Prototype One

This was the first prototype we produced and tested with Children’s Museum staff and visitors. Our intent with this prototype was to get a sense of the scale of our play area, experiment with different materials, and explore different technological directions. It consisted of a yoga mat and memory foam pad for the base, pool noodles fixed to a wooden stand for the grass, and two different methods of pressure-sensitive sound, detailed below.

Doll Audio Modules

These little modules would play back a small recording when pressed. Their main advantages were their cost, simplicity of use, and the ability to overwrite their sound clips with new ones at any time. However, the sound they produced lacked volume and we could not place them under the surface of the installation without them making a bump in the fabric. It was clear that while these were great for early prototypes, a replacement would need to be found quickly.

Pressure Sensors

The cloth pressure sensors we fabricated proved valuable due to their small physical profile and general-purpose usefulness as a switch. To start, pieces of conductive cloth were adhered to two pieces of nonconductive cloth. Sandwiched between these was one circular piece of Velostat, a conductive plastic that changes electrical resistance with the application of pressure. Once assembled, the cloth pressure sensor could be hooked up to an Arduino as shown in the attached diagram.

Sewn Speakers and Pressure Sensors

One idea for audio playback that we wanted to try out was embroidering fabric speakers into the base of the installation using conductive thread. When current was passed through the thread and a magnet was held near the surface, sound would be produced. Our cloth pressure sensors were first wired to an Arduino. A Processing script serially connected to the Arduino played sound when the sensor was stepped on. We now had a setup without anything kids could trip on, and as a bonus the speakers formed a very cool pattern. However, we hit a wall with sound quality again – the speakers simply were not loud enough to be noticed in a museum environment.

External Speakers and Pressure Sensors

Since we were trying to explore as many audio setups as possible, we thought it sensible to also try playing audio through conventional speakers instead of through the cloth speakers. It’s often been the case that the simplest solutions provide the best compromise between pragmatic needs and aesthetic desires.

    What Worked

  • +
  • The pressure sensors were responsive and could easily read differences in the pressure being applied.

  • +
  • The grass setup was reliable and structurally sound.

  • +
  • The build-a-bear modules proved great for rapid deployment due to their simple mechanics and cheap cost.

  • +
  • The cloth speakers provided a great aesthetic addition to the installation.

    What Didn't Work

  • +
  • The memory foam pad was too soft and comfortable. Children wanted to lie on it to relax, not perform physical activity.

  • +
  • The cloth speakers and build-a-bear modules were simply too low in volume to be useable in an environment like the Children’s Museum.

Prototype Two

During the testing of the first prototype we received a lot of valuable feedback from staff and visitors. We learned of several ways we could improve the mechanical and structural aspects of our prototype. For the next iteration, we settled on using conventional speakers in concert with the pressure sensors, and added an ambient noise track. We discarded the memory foam pad, as we felt it discouraged more active physical play. The entire open area was now covered with green cloth, underneath which we added a dome from a children’s turtle sandbox to create a small mound. In addition to the grass we added a rock pillow embedded with an accelerometer which would play sounds when shaken. The grass now had an embedded tilt sensor so that it too could play sounds when sufficiently moved. We had the privilege of presenting this to Eric Carle’s representatives during their visit to the Pittsburgh Children’s Museum on October 28th, 2014.

External Speakers and Pressure Sensors

We chose to carry over this feature from the first prototype because it solved our playback volume issues. Now we needed to make the speakers less conspicuous. By placing the speakers underneath a plastic dome under the base cloth of the play area, we could hide them from view. The dome also gave the installation a pleasant little hill, which added to the pastoral look we were trying to achieve.

Tilt Sensor

We wanted our giant blades of grass to make noise when jostled by kids, and a simple embedded tilt sensor proved a perfect trigger. Two balls in the sensor complete a circuit when the sensor is upright, and break the circuit by rolling away when the sensor is moved. The tilt sensor thus acts as a simple switch for determining orientation, allowing us to play a sound when a change in orientation is detected.


One idea we had was to hook accelerometers up to the stuffed “rocks” strewn bout. We could read information on which way an accelerometer was facing in the form of a vector. By periodically reading in these vectors and taking their cross products, we could derive the value in degrees of the accelerometer’s change in direction over a given time period. If the change was large enough, we could then play “rock sounds” over a speaker nearby. The major limitation of this idea was the necessity of a wired connection running from the rock to the Arduino, which would severely constrain the rock’s range of movement. Furthermore, this setup easily ran the risk of damage or disconnection due to over-enthusiastic kids throwing the rocks around. It became readily apparent that an analog solution of rattles embedded in the rocks would work as a better alternative.

    What Worked

  • +
  • The addition of a mound in the middle really helped with establishing the pastoral look we were trying to achieve.

  • +
  • The updated grass and rock proved reliable and fun sound-makers.

    What Didn't Work

  • +
  • The accelerometer in the rock required a wired connection, which reduced its mobility and structural integrity too much to be useful in the long run.

Our Final Prototype

To view our final prototype in detail, please see our homepage.