Electronic Texture

We can send text, video and audio over the internet, why not textures?

Disclaimer: This was the culmination of a series of thoughts in 2016 that led me to think about what it would take to send textures over the internet. What is written is not comprehensive, backed by evidence, or even true. Yet, it is left here as a testament to my willingness to explore the ideas in my head no matter how outlandish they may be. I also like the artwork, and it would be a shame for it to go to waste!

We can electronically send sounds, words, images, videos, and many more types of information. Combinations of these data types can create a convincing reality whether it is simple a video or a whole world to be experienced through VR, but there are still many forms of data that we have yet to be able to send electronically such as smell and texture. The problem is that we do not know how to convert these ideas into electrical signals that are easily transferable between devices and then reconverted back using hardware.

This  problem may stem from a variety of reasons. Using scent as an example, the way in which our body detects smells is still quite a mystery to us. There are two predominant theories, but that is not even the problem. Even if we did understand how we smell, that does not assure our ability to create any smell using a simple list of resources. We would require thousands of ingredients to replicate any given smell effectively. It is a major obstacle to overcome, but there is another data type, mentioned above, that could be possible to replicate: Texture.

The first question that must be asked and answered is “What is texture?” The most important aspect of a given texture is not its composition, but its structure. What do I mean by that? I mean that the tiny ridges and bumps that are found on an object are the most important aspect to how we feel something. Another important factor is the way in which a material or object responds to the force we exert on it. For example, if we touch a cotton ball it will absorb and dissipate the force applied thus allowing it be very “squishy”, but a piece of wood when pushed upon will not give way thus making it “hard”.

There are more important factors such as local randomness vs. uniformity. What this means is whether the object has continuous features such as marble being consistently smooth, or whether its surface randomly rises and deepens, such as a rock found outside that is jagged in some parts, smooth in others. Another important factor is actually how a material chemically interacts with our skin. When we feel a liquid such as an acid, it burns which can arguably be counted in the texture of that liquid. Another example is that water loosely creates bonds with our skin so is very easy to remove, but honey sticks onto itself and our skin, so that when we take it off with a towel we are left with a “sticky” texture. The most obvious example is also temperature. Another possible factor is the visual aid that helps contribute towards texture. This does not belong in the physical aspects of texture, but our expectations of a texture can influence how we feel it, but this can be separated from physical texture and named visual texture.

So, the factors we have when looking at texture involve:

  1. Microstructure
    1. Local homogeneity vs randomness
  2. Response to our input
  3. Chemical interaction
    1. Temperature
    2. Reactivity

This list is by no means exhaustive as I am sure I am missing numerous critical points, but it is a good starting point for the goal I have in mind.

If we wanted to be able to electronically send texture we would need to figure out how to convert these factors into a quantitative form, and figure out a method that we can use to recreate our quantitative values back into the original texture. This is quite difficult, but to help us off, we can discuss how the hardware would work to be able to recreate these factors in any given scenario. The method I thought of involves a massive amount of tiny pistons that would allow us to recreate a given microstructure. We would be able to give each piston a value that would tell it to go up or down by a given amount. This would create convincing ridges and bumps that would accurately represent the terrain of whatever sample we intended to recreate. On top of each of the given values to the pistons, we would also need to give it a leeway that allow for a certain amount of “flex” when applied pressure on. If we wanted to imitate cotton, the piston would be very easy to push down and would go back to their original position over a given time and acceleration, but wood would not allow the pistons to move whatsoever unless a massive amount of force was applied.

This could be achieved by giving the piston a position value say from 0 – 100 (arbitrary) as well as a force value, say from 0 – 100 (arbitrary) that determines how much force be applied until the piston starts to move, this force value would corresponds to how rigid the texture is. And finally we would need to provide an acceleration graph that helps the piston to understand at what rate it goes down when the force value is met, as well as what rate it goes back to its original position when the force returns to zero. This would imitate rigidness, elasticity, reactiveness, sensitivity, softness, among many other words we associate with touch. Again, this is simply an idea on how it will work and this idea may be utterly insufficient in what it aims to do.

Another idea on aiding the pistons, or increasing their effectiveness, it to place a bud on top of the piston that is filled with fluid oil and stimulate it with electricity to create sensations and  even smaller patterns that help mimic micro patterns and sensitivity.

An additional method that could be employed to make the project more feasible is introducing an axis of rotation at the base of each piston. This would allow them to be flexible, and it would make recreating certain textures much easier. It would involve creating a whole new set of variables such as the force value for rotation, the rotational acceleration graph, and more.

Finally, we approach the problem of chemical interaction. I believe for early editions this would have to be ignored for the most part as it runs into the same problems at trying to imitate smell: you would need extensive resources to do it effectively. Oil filled tips stimulated with electricity may be able to simulate certain sensations, but it would have severe limitations.

The next step in creating transferable experiences most likely lies in quantifying and reproducing texture. The methods and details will be hard to work out, but it is viable.

Leave a Reply