Archive for: March 2008

There are several new digital drawing and painting tools I’d really like to see made available, based on the behaviour of real paint.

The first of these is pretty simple. Today’s invention is a ‘bucketfill’ tool whose paint progresses radially outward from the point clicked on the screen at a pace which allows the user still to keep track of the paint’s leading edge -and stop its progress if necessary.

Not only would this be more like dropping a dollop of real paint (you could incorporate various realistic sound effects and transient translucency as well -if you had an embarrassing surplus of computing power). I’d also include some surface tension, viscosity and wave physics so that, as the paint front approached any small inadvertent gaps in the surrounding ‘dam,’ it wouldn’t necessarily spill out and instantly colourise the entire screen but instead highlight these glitches so that they could be plugged.

Aside from these aesthetics, the time course of filling a particular shape (from a variety of known starting points) could be used by the program to identify the shape being painted.

It’s often very annoying that I can be doing my best impression of exercise when the music on my MP3 device starts playing at completely the wrong beat.

Today’s invention is therefore a mechanism on a portable music player which will allow selection of a particular beat and play only tunes with that rhythm (or one close to it).

This could also be interfaced to a rowing machine or treadmill so as to select tunes that corresponded to the current stroke rate or pace. Ideally this would be set to choose a beat which was slightly faster than the current exercise frequency in order to keep up one’s work rate.

I’m not sure exactly how sensitive to rhythm variations we are, but it might also be possible to play any given piece of music at a slightly elevated rate, in order to better match the target level of effort.

Finite element methods are commonly used in engineering design. You create a CAD drawing of your component, automatically divide it up into a mesh of a million tetrahedra, apply some realistic loading boundary conditions and material properties and finally compute a detailed stress map within the design. If you are nearing yield stress anywhere, it’s time to think again either about geometry or material spec.

Today’s invention is simply then to use the 3-D mesh of tetrahedra as a specification from which to actually build your design. I’d like to see this happen using a rapid prototyping system. Each tetra could be ‘printed’ as a net of four triangles, joined along three edges. As each is produced, it would be automatically folded and the six free edges bonded into the 3-D ‘pyramid’. The system would spit these physical elements out in the correct sequence (and orientation) for them to be bonded together to create the overall design.

This would already have been shown to withstand the required in-service stresses. It would represent a huge material saving, compared to using bulk material, and although the final shape would still have a rougher ‘crystalline’ outer surface, this would be a structure with low form drag and a genuine wow factor in the marketplace (think Ducati frame taken to extremes).

Search has become fascinating for two reasons: the recent availability of an abundance of interesting stuff (if only you knew where it was) and the possibility of automating its retrieval.

One of the biggest problems, it seems to me, is that we currently aren’t really making use of all the tools available to us. It’s still hard work to find things visually or kinaesthetically in a reliable way, but sound offers a more tractable approach.

It seems that social networks are ‘small-worlds’ -everyone is only six handshakes away from the president because these networks consist of clusters of individuals and quasi-random, long-range links between clusters. These arrangements form because people tend to get introduced to others who are slightly better connected than average.

Small-world networks (and the Internet is one such) have a major problem which is that they can be seriously damaged by the removal of a small number of key people at the centres of a smallish number of clusters. If their identities are known, (and those people are by definition, self promoters) the whole system is at risk.

Today’s invention is therefore a ‘wallflower’ network. One in which newcomers are naturally linked to people who are less well-connected than average. This results in networks which consist of loops surrounding voids…a bit like the structure of lace or froth.

I haven’t yet worked up a simulation, but I believe this means that you get a similar level of connectivity, without the vulnerability -everyone is president in such a network.

Last night, my spam filter detected and removed 779 items. Although I never check the details of what’s screened out these days (I might be bulk deleting some real messages) it seems to be pretty effective in eradicating anything to do with investment scams or phishing ploys. Amazingly, the spam that does get through almost always contains some kind of (pharmaceutical) brand name.

Today’s invention is simply to apply a crude, automated translation program, such as Bablefish. This will come out with results like “xanax nu koopt de bevordering.” Any such system will fail to translate brand names and so messages attempting to sell you things will leave words unchanged after translation. The presence of untranslated words is thus an indicator of unwanted marketing.

Using different translators, across a sequential spectrum of different languages, will enable a higher level of reliability.

Harvesting trees, in even a responsibly managed plantation, is a messy and energy sapping process. Even when they have been cut down, the forest floor is still littered with roots, making it harder to replant.

Today’s invention is a way to allow the trees to grow naturally, whilst making their harvesting more straightforward.

Each seedling is planted in a thin layer of soil under which has been buried a small, square flagstone. The stones are impenetrable to tree roots and they interlock so as to form an effectively continuous surface.

The trees grow roots laterally and within a limited depth of earth. At harvest time, a root cutting saw can be guided along the edges of the square stones (perhaps mounted on a small robotic carriage). This will so destabilise the trees that a moderate wind will then be enough to fell them (pulling up most of the remaining roots at the same time).

Placement of the seedling stem to one side of each stone, when planting, will allow the direction of its subsequent fall to be predetermined.

Disliking tattoos as I do, I’m surprised to find myself suggesting today’s invention.

It’s simply tattoos which are drawn (inscribed?) as a stereo pair. Each image would be of some favoured 3-D object seen from a slightly different angle, so that people could ‘fuse’ the 2 images and see eg a 3-masted schooner sailing across somebody’s backside.

For those people who haven’t yet learned how to fuse two images, by slightly crossing their eyes, it would be possible to tattoo one image in red, one in blue and then view them through a pair of colour-filtered spectacles.

Certainly a talking point at beach parties, athletic meets, biker rallies and high-security prisons.

It’s not like I’m tight with money (honestly!) but waste bugs me. (Actually the biggest waste is of human creativity, which is often devalued because it challenges those without it so effectively -rant ends, or at least pauses).

Anyway, today’s invention is yet another way to get at the dregs of something valuable and viscous, rather than pitching it out with the packaging.

A container is made of flexible plastic, in the shape of a cylinder, as shown. It would have a screw-off lid as usual and when this was removed, contents could be extracted by driving a rod (supplied with a hook at the end) upwards against the base, inverting the cylinder.

When enough product has been extracted, at a given sitting, the hook on the rod would engage with a simple bar inside the base and allow it to return the cylinder to normal shape, by dragging it backwards.

This avoids any contact with the content, keeps whatever it is reasonably well mixed and eradicates the possibility that anything remains in the package because it can’t be accessed.

Who can be bothered sorting clothes before piling them into a washing machine? Actually I learned this lazy habit as a student, which accounts for the fact that everything I wore eventually took on a neutral khaki-grey hue (fine if you’re an Engineer, not so good at art school). Nowadays it’s just easier to wear navy blue exclusively.

For people who aren’t rod monochromats and who inexplicably like to wear delicate, coloured clothing, shouldn’t the washing machine be capable of working all this out for itself?

Today’s invention is a simple barcode reader attached to such a machine. Each item has its code scanned as it is passed through the door (it might require that anything you care about be equipped, post-purchase, with an iron-on patch, since RFID doesn’t seem as if it will be available any time soon).

The scanner allows the machine to issue a verbal warning before the item has entered the machine saying eg

“It seems that you are about to add a bright red woollen jersey to a whites load…but this is the fourteenth time this item has been washed, so it should be ok…proceed?”