Good afternoon ladies and gentlemen my name is Lionel Dean. I am a practicing product designer, I have my own practice, I am also a researcher – I am studying my PhD at the moment, and to some extent I am an artist, as well. Back in 2002, I had the opportunity to do a one year design residency at the University of Huddersfield. I wanted to use the opportunity to look at what might be done with rapid prototyping techniques. Obviously in my design business I had come across rapid prototyping and the use in the design process. But I wanted to see what could possibly be done with this, what the future of this could possibly be. Whether you could use this for production for rapid manufacture and if you could use it for rapid manufacture whether you could use this to individualise products in some way.
With rapid prototyping techniques the cost is based on the size of the model – on the height of the build chamber – and it doesn’t cost you any more to produce two things that are similar in size but different shapes than it does two that are exactly the same shape. So the economics of mass-production don’t apply here. So could you produce an infinite stream of products that were subtly different in the same way that you probably would historically as a craft maker making things by hand. Things that vary slightly, might vary a little bit with the material you are using or you might feel a little bit differently and tweak something on the day and somehow there’d be a little bit more of a human element to the product, there would be a bit more character of their own. Perhaps this would be something you would want to treasure and not throw away because there wouldn’t be another one quite like it. It would have some sort of personal value to you.
How this would be done would be using a combination of rapid prototyping and parametric CAD. Parametric CAD is just CAD where objects are defined by relationships between different values rather than absolute values. So you have a model there and if you change the length the whole model will update to accommodate that. So you can have a situation where this model here [slide] - each one of these tuber forms is defined by a series of cross-section circles and a skin is then generated between them. Then you can tweak these circles, you can twist them, scale them, translate them and the whole for will update accordingly.
You could have a situation where you have a factory or a production line and there is a designer sat at the end at his CAD screen tweaking the model each time. But that would not be an automated process – it would fall down – because you are relying on the skill of the craftsperson still to be sat there with his computer. I wanted to try and automate this system. I wanted the designer to be able to define a set of rules for this object so you almost choreograph this like a movie. You then set it going and then it carries on in its own sweet way generating design after design. But each one, hopefully with the rules that you have set is still true to the design direction you wanted to give this product in the first place.
When I started off I thought this was a blue skies research project (back in 2002) then in 2003 Materialise showed their first collection. This obviously pointed the way that this actually wasn’t so far away in terms of the rapid manufacture being already viable as a production process – albeit yes it is expensive to design objects but in reality the prices of the lamps that Materialise sell aren’t too different from the high end Italian manufacturers: Artemedia, Flos – they are comparable in terms of price. So this is a viable process already and that made me rethink slightly how far away this was and maybe it wasn’t a blue skies project it was something I could start doing straight away.
While I was generating design outputs for the project I also spoke with Materialise and we started working on a couple of projects and we productionised a couple of designs. The first one was RGB which is based on the tuber forms - this is it in laser sintered Nylon. They wouldn’t think about individualisation just yet I’m still working on a company to work with on the idea of individualisation there are a lot of hurdles to that which I will come to. With this particular one the software was there to generate the form we produced – four different ones, but four is as far as they want to go for the time being.
This is Creepers and the idea of this one is that it is a series of petal forms and the petals catch the light from some very tiny LEDs just 5 mil LEDs. The idea is this creates a space divider or room divider with light. Really the idea here was the idea of generative design and forms evolving to create this pattern making the leaves different with each cluster.
This year I worked with an Italian manufacturer Kundalini and we produced this table lamp for them - Entropia. This is quite a leap forward in terms of the industry because here is a company that have nothing to do with rapid prototyping. Materialise started off because their lamp division is actually a very tiny division of a huge company that markets software and is probably I would think the largest manufacturing bureau in Europe. For them it is a little bit of advertising and at the same time they have a vested interest in the technology that they are trying to promote. With Kundalini here was a company coming to rapid prototyping for the first time they hadn’t used it in development before. They were purely interested in what they could do with this in terms of form. And what they wanted to achieve was a form that would baffle people - that people would have no idea how to go about manufacturing something like this. They didn’t want any rules, patterns or repeats that anyone could identify. At the same time, as a slight contradiction to that they wanted it to be obvious that there was some process behind it – they didn’t want something that was just random because they didn’t think that would have a perceived value. So they wanted some evidence of process but you not to be able to figure out how that worked. It was a tricky brief to deal with in that respect. But what I wanted to do and what I ended up doing was applying the rules and relationships that I used to generate the FutureFactories collection but rather than changing the overall form I was changing components within that form. So it is a tree or bush-like form made up of a number of different components the circular ones I thought of a flowers and the rest are leaves. The flowers have a hole in the middle and the stems curl back behind that flower to block up the hole so you can’t see directly through to see the light behind. But there is a lot of room to manoeuvre with that form in terms of how it can change. So I think that the flowers – there are around two hundred in the form and every one is different. In the whole form there is something like a hundred chains of components - the chains are about twenty elements long. There are a lot of different components to the design but we set up a number of rules so each of the components would change every time and then we used that to generate each different one as we applied it.
This brings me to the project that I’ve done for this exhibition I wanted to use the opportunity to work on something on a slightly larger scale. And also I had been thinking for a while about the notion of using the potential of reverse manufacture – designers taking on other people’s designs and adding bits to them or if you had an artefact at home maybe instead of chucking the thing away when it breaks maybe it can be modified in the future so you might have something that gets renovated rather like architecture where you get bits added to it from different periods to an object. There are also very practical issues it is a very large piece to have manufactured so I just wanted to stick with just the back and the arms. The idea was to take an iconic chair – Starck’s Louis Ghost chair fitted the bill quite well in terms of it being a chair that a lot of people would recognise and being transparent it was a nice sort of plinth that didn’t dominate the rest of the work. The idea was to chop the back and arms off and replace those. The back and arms would just squeeze into the largest laser sintering machine that is available at the moment. There are larger SLA® machines – you might have come across the work of Patrick Jouin who has done a chair with Materialise which is longer in terms of the bed length but it is SLA® and it is functionally not as strong. It is very much an art exhibit as opposed to something that is functional.
So, off go the back and arms and the one you’ll see down in the exhibition is this one here. The back form is taken from the idea of button leather and each one of these ‘buttons’ floats independently. The arcing ribs you see across the back all act as springs because the Nylon has flexibility and so they’ll all float independently. Rather like a spring mattress you lean back into this and it adjusts to fit your back. This is where I come into the argument about random design and do I do random design. My argument is that I don’t produce random designs I try and get generative software to produce iterations of my designs but I don’t feel they are random. I think ordinarily when you do a design process like this when I had this idea in my head and started sketching this thing out there is a point where you say to yourself "how many buttons?" I think I started off with twenty two. I think that changed with the first version to twenty four. But that is not really fixed it is slightly arbitrary. With the generative software what I’ve ended up doing is giving it a window – I think that with the version we have now it is anywhere between twenty two and twenty eight. Also, the position of those isn’t terribly important. First of all, I started off positioning them at random and then spreading the forms out and seeing how they looked. Again, the positioning can be part of the programme.
I’m working with a software development tool which simply gives you a nice little halfway house between programming and something that is a little easier to work with as a lay person. What it is doing to start off with is positioning the buttons at random. First of all it has already decided how many buttons there are going to be and it is now trying to position them so it is dropping them into place at random within the envelope of the back. If it clashes with one of them it removes it and moves it to another position. It will do that until it has the right number of buttons and then they’ll begin expanding.
One of the problems of working with rapid prototyping – I thought when I’d moved to rapid prototyping that it would be great you’d have bureaus produce endless samples for you and you would be able to do this as an iterative process. But in fact the cost of the process means that very much you end up doing this in one hit. With this chair we had structural implications with those springs in the back to work out how stiff they needed to be. The whole thing had to be done pretty much in one hit. We did talk with the University about doing this with Finite Element Analysis to try and work out the stresses. But it is such a complex problem in that the way you sit in this your weight is spread across the entire back of the chair and it is a very difficult problem to work with. Fortunately, it came out first hit.
It is very different to how things used to be where normally I’m working on a design project I had to have a studio full of bits and bobs and models at various stages of this project - physical models – whereas now I’m seeing the thing on the screen the whole time. In fact with Entropia the prototype was built out in Italy I didn’t see the models until long after the client had seen them. Which is quite an eerie feeling when the client – you are talking to the client on the telephone – he’s looking at the model and you are not you’re seeing it on the screen but it’s not quite the same thing.
So this is the model as it ended up. It is quite an interesting to work with something other than a lamp form. SLS® is – if you’ve seen the process – it is a textured finish that you get with this. It’s fairly rough and ready and interestingly enough I was at a conference last week the TCT (Time Compression Technologies Conference) where they were talking about the future of rapid manufacture and saying that the big hurdle is the finish of SLS®. I was saying in my talk how beautiful the finish of this was because as a designer working with objects that are on the art fringe it is a very beautiful material. Particularly when you use it for lighting because when it is back lit it is almost like a wood grain finish. The problem with it is it being a textured finish it picks up grease basically so if you handle these things they do get dirty over time. Also they are UV sensitive so if they are in strong sunlight they will yellow over time. I think there are a lot of people out there with nice expensive lights that maybe in ten years time will be a bit sort of nicotine yellow.
Finish is one of the big issues it would be nice to have a really nice glossy finish you could achieve via rapid prototyping. The first objects I did were hand finished that was necessary because I was using some of the cheaper 3D printing processes. So to get the exhibition quality I had to use hand work but really then the whole model falls down if the idea is the designer sets up the template then manufactures straight from the computer. If you then take them back and do a whole lot of hand work on them the whole model falls down. So that had to be avoided. Then with the raised profile of the project and the backing that that brought it was possible to use some of the more exotic processes and so went into SLS® Nylon and you can see in the close-up there the texture that is coming through from the rapid prototyping. Which in lighting is a very beautiful finish but not quite so applicable when you see it in an object like a chair. That is a close-up of Entropia again you can see the striations the lines from the process. And that is where I am at the moment.