Article

Production line of robots assembling each other4 comments

The Next Big Thing

Published at 8:07pm on 05 Mar 2007

In the last 25 years we have seen a number of key technologies emerge that have revolutionised how we work, play and interact with each other. Could these innovations be part of a grander design? And if so, what's the next piece in the puzzle...?

Some would say it's a bad time to be a futurologist.

For many, the turn of the millennium has been a big disappointment. There are no flying cars, no men on Mars, no base on the moon. Our every need is not tended by robot servants, we haven't cured Cancer or AIDs and "virtual reality" video games still suck.

None of the great technological miracles we were anticipating during the latter half of the 20th century have come to pass, and yet, in the last quarter of a century we have seen four major technology revolutions, none of which were really predicted by so-called futurology experts, or even science fiction:

  • The personal computer

    In 1943, Thomas Watson (Chairman of IBM) predicted that the world market for computer technology was maybe five mainframes. 50 years later there are more computers on Earth than there are people, and in addition to the obvious desktop and laptop machines we have PDAs, set-top boxes, games consoles, and any number of other examples of other near-general purpose computers working their way into consumer devices.

    Even forward-looking science fiction of the late 70s (Star Trek, Alien, 2001) expected that a computer was a device that one might find one of on a starship. Ken Olson, Chairman and founder of Digital Equipment Corporation said in 1977 that there is no reason why anyone would want a computer in the home. Fast forward to the present day and a typical teenager might well have two or three on their person at ay given point in time.

    Computers have of course completely revolutionised our lives. They act as surrogate calendars, calculators, secretaries, assistants, sweatshops, televisions, artists, special effects houses - you name it. Computers are doing for the mind what industrial tools did for the body - taking over the boring repetitive tasks and leaving humans free to focus on what we're built for - the application of creativity.

  • The Internet

    Remember the 80s? We had computers and we thought we were pretty smart. We could imagine that from computer technology would come robots and automated factories. We would replace menial jobs with computers, and retrain everyone as programmers. Computers would be our labourers, our soldiers, our ticket to the stars, and perhaps even one day our friends (AI - another technology that's been slower to develop than we expected).

    But telephones? Ham radios? letters? Shops? Nobody would have expected that computers would do away with these. If you want to send someone a message, or provide customer services, or go shopping, then what has a computer got to do with it?

    Nobody really saw the Internet coming. Even when ARPANET went global, and UNIX nerds were happily playing networked text-only games of Dungeons and Dragons via a modem, or sending each other porn rendered in ASCII art, nobody seemed to see the huge revolution that was about to hit.

    Thanks to the Internet, an idea can now be propagated instantaneously by anyone, with its success based purely on merit or interest rather than that individual's affiliation with the local radio or TV stations. A savvy blogger can spread a scandal around the world before a reporter even makes it to the scene, and the Internet community can fact-check and correct a media story within an hour of its release.

    Retail too has been revolutionised. Software or other products can be created and distributed by anyone in the space of minutes, and without having to spend money on getting it shipped out to retailers. Even conventional retailers can now use the Web to massively slash distribution costs, shipping products from a remote warehouse to the customer's door without any need for a physical "shop".

    The Internet has changed our perception not only of communication and media, but of information as a whole. One way or another the Internet is eventually going to make all information known to mankind available instantly and free of charge to anyone on earth.

  • The mobile phone

    It's not really a big idea or a particularly new one. Mobile phones have been around in one form or another since the late 70s. Okay, so they were bulky and expensive, but so is every technology at first. A few people probably got very rich by investing at the right time, but most people simply never saw that the mobile phone was going to go from business executive's toy to must-have consumer gadget within the space of a few months at the end of the 90s.

    A decade ago, you got strange looks if you used a cell phone on the train. Five years later everyone above the age of 10 had one. Within the space of a couple of years, busses, hospitals, cinemas and lecture halls went from being completely unaware that cell phones existed to being plastered with signs saying to switch them off or keep the volume low.

    Old-age pensioners, who would have sat stony faced on the bus, glaring at cell phone users whenever their ring tones broke the tranquility just a few years ago, now spend the entire journey gabbling away to their grandchildren.

    I can't imagine how people used to meet up in unfamiliar locations, or how anxious parents used to keep tabs on their children venturing out with friends one their own. I know that we did it - if I'm honest I even remember how, although it seems absurd now to think that I used to stand in the freezing cold on street corners waiting for my friends, or that parents used to have to contemplate calling the police if their kids were an hour late getting home, instead of just calling the kids!

    When exactly was the point at which we all went from thinking that being able to contact anyone at any time without knowing where they are was a frivolous luxury, instead of being an obvious and necessary requirement of modern living. And why didn't it happen sooner?

  • The MP3 player

    The MP3 player (or rather the iPod) is changing the way people purchase and listen to music.

    It's not merely about the ability to play tunes on the move - Sony achieved that 20 years earlier with the Walkman, and whilst this was certainly an impressive new technology at the time, it can hardly be called a revolution. The key feature about the MP3 player is not that it plays portable media, it's that it plays disembodied media - MP3s are the first distributable media format that requires no physical medium to transport it.

    In one fell swoop the success of portable MP3 players has marked the beginning of the end for the Walkman, the Minidisk, the CD, and quite possibly the entire music distribution industry if they continue to mismanage the issues of digital rights in the way they have been.

    MP3s, and other formats like AAC and Ogg are slowly wiping out the need for physical media formats, and in the process they've shattered any hopes that distributors had of continuing to profit indefinitely by redistributing existing works in new formats without adding any value or creativity.

    Expect the same thing to happen with video in the next 5 to 10 years as bandwidth catches up. I predict that the HD-DVD/Blu-Ray will be the last physical video formats, and even those are unlikely to enjoy the success of their precursors - the video cassette, and the DVD.

The connection

At face value, these listed technologies seem largely disconnected, but I believe there is a fundamental connection between them. There's some crossover between actual devices of course (MP3-phones, Internet phones, PDAs, etc) but I'm not talking about that, I'm referring to a guiding principle that underlies the success of these technologies, and from the perspective of which we can say that all these things were inevitable.

I believe that this key combination of personal computation resources, global networking, wireless communication and freely transferrable information and media are all building blocks in a grander scheme. On their own each of them fulfills a useful short-term consumer goal, but together they become much greater than the sum of their parts.

And I have a good idea what the final piece might be.

At the beginning of this article I said it was a bad time to be a futurologist. Futurology is a fools game. For most of history a futurologist could have said that tomorrow will be exactly like today, and he would have been right. Nowadays, trying to predict the direction that technology will take is essentially impossible. The philosopher Karl Popper pointed out that one cannot predict the future growth of knowledge since predicting knowledge is essentially the same as possessing it.

And it's not enough to predict the actual advancements of technology, though that in itself is essentially impossible to predict unless you are guessing at trivial extensions of existing trends (the most famous example being Moores law). You also have to predict what the public will go for. The revolution point in computers or mobile phones wasn't a purely technological leap - it was a social one - the point at which the public decided that the benefits of a technology (whether the benefits lie in utility, productivity, or even fashionability) outweighed the cost.

It's not sufficient (or perhaps even necessary) for technology to reach a particular apex of price or form factor - the portable electric trouser press is probably a pretty well refined technology, but do you know anybody who has bought one? What about the electric car? In truth it is the public that will decide when they want to go out and buy these products en-masse, and once that happens the resultant injection of funding will usually bring about any necessary technological advances to keep them buying. Supply follows demand.

It is with some trepidation then that I put forth a theory for what technology will be the next to go from an expensive fringe toy to being so ubiquitous that our children will simply not understand how the world functioned before it existed...

The problem

Every solution requires a problem to solve. The current model for manufacturing is that companies design, build and supply products to consumers. Consumers do not buy the exact products they need or want because they don't exist - they buy the ones that most closely match their requirements.

For many consumers this is good enough. The average consumer in fact does not have a very good idea what they want, or even what is possible and so these companies provide a valuable service by designing products and then advertising them - they inform the consumer of what is possible, and the consumer can then accept or reject it.

This is both a desirable and necessary model, but it has limitations. The design and manufacturing industry is like a benevolent dictatorship - most of the time is knows best and does what the majority of people want. In instances where it doesn't, people must weigh the cost of oppression versus revolution, and usually they decide that the oppression is not worth fighting over. This is why companies like Microsoft (and indeed Apple) are so successful - even when the majority are dissatisfied, the technological ostracism of going it alone and joining a rebel faction (Linux, in the case of the example) is too much to bear for most people.

People have traditionally endured unsatisfactory products and services because the barriers to using or supplying their own alternatives are too great. And yet in the world of software we are witnessing a gradual change...

Open source software has created a new design and distribution model for software. It has made it reasonably feasible for people to start providing their own software solutions by removing the barriers of complexity and cost. Even with the Internet, it's impossible to find enough of the right kind of people to work on a big, closed-source system because creative people have different visions and without paying them to keep them on track they will naturally want to go their own way. The creative commons licenses ensure that they can do this without taking their code with them and preventing others taking up where they left off.

Open source has made it viable for software to branch into tens or hundreds of variants to suit every single user. Even with a common source tree, the plug-in mechanisms in programs like Firefox mean that every user can configure their copy exactly how they want it, with far more variety than any single programmer could anticipate or code for.

With hardware though, we are still stuck with whatever the industry is prepared to give us, and industry agendas (such as protecting monopolies over digital media distribution) mean that the gap between what we want and what we get is growing by the day.

The solution

We are reaching a point now where it is becoming technically and financially feasible for consumers to have hardware designed and manufactured in a bespoke fashion. The actual technology behind this is not so important as the concept; there are many competing technologies working to bring this about, and I expect them to become commercially and technologically viable in roughly the following order:

  • Bespoke manufacturing

    The least "revolutionary" of these technologies, but currently the most practical. Bespoke manufacturing means a company or companies who are set up to do very small production runs of products at a reasonable cost.

    Traditional manufacturing models dictate that the price of manufacture goes down with increased size of production run. Factories are set up in such a way that tooling (the creation of customised moulds and templates for a specific product) is very expensive, but the cost is amortised by re-using that tool for hundreds or thousands of copies of a product.

    It is prohibitively expensive to manufacture one of something in a traditional factory, so very few will even accept orders for it.

    A bespoke manufacturer works to a different model. The tools used are all multipurpose, cutting the cost of tooling to essentially zero, but resultantly increasing the cost per unit. For traditional manufacture this makes no sense, but for bespoke manufacturing it actually works out as worthwhile because the increased cost of each product may be relatively small compared to the reduction in tooling cost. If a big factory could cut the cost per product from, say $5 to 50 cents then they wouldn't mind if the extra tooling cost $10,000 as long as they were planning to make a hundred thousand units. If you just want to make one of something though, you'd rather pay $5 than $10,000.50!

    As long as manufacturers only cater to large production runs, designing and manufacturing your own hardware is economically unviable. Another problem is that providing designs in a suitable format is not trivial for a home designer. The industry standard software for industrial design is complex and expensive.

    Fortunately some companies are emerging that are prepared to handle one-off production runs for consumers: Wired ran an article on Squid labs' eMachineShop - essentially a consumer software package for doing CAD. There's nothing remarkable about this in particular, except that once you've made your design using their software, Squid labs will build it for you and then deliver the finished product to your front door.

    Bespoke manufacturing as a service is all very well, but it still requires the services of a 3rd party, and it still means waiting for the finished product to ship.

    Most customers are not actually going to want to design their own hardware - at least not at first. Rather, they will want to mix and match designs made by others, perhaps tweaking the colour or accessories. In this instance bespoke manufacturers do not offer such an obvious benefit over what traditional industries can achieve by providing basic customisation features. Apple did this in the late 90s by offering iMacs in various colours, and Nike currently offer a couple of products with customisable designs that you can tweak via a Web interface before purchase.

    There are problems with this approach, notably warehousing issues. If you offer five flavours of a product that is manufactured by traditional methods that means storing five different types in your warehouse. You then run into a situation where a customer wants a lime green iMac, but you've sold out of those and can only offer them a strawberry red model. The more options you have the worse this problem gets, and unless demand is relatively equal for each product variant it becomes a logistical nightmare to ensure that supply matches demand.

    Bespoke manufacturing provides some advantage here, but perhaps not really enough to start a revolution. For that to happen we need something else...

  • 3D printers

    The basic concept of the 3D printer is pretty simple. A 3D printer is to a manufacturing plant as a regular printer is to a printing press.

    The dream is that one day rather than sending hardware products to consumers via the postal service, schematics can be transmitted electronically and people will simply "print out" a copy of whatever they have purchased, whether it be a table lamp, an iPod, or even a better model of 3D printer. You've heard of downloading software or firmware updates - how about downloading a hardware update!

    The reality is a bit less exciting - at least for the time being. 3D printers do exist; They are used primarily for rapid prototyping - basically a process where a foam or plastic mockup of a design can be quickly constructed so that hardware developers can have a tangible sample of their product to play with before committing to spending millions on tooling so they can manufacture it with more durable materials.

    These printers generally work in one of two ways: Either they build up an object layer-by-layer by depositing some form of resin - essentially bonding multiple "2D" printouts together to form a 3D whole. Or they carve a shape out from a solid block of material, a process called milling.PhysicsWeb recently posted a story about a rapid prototyping machine that uses a third method - it folds shapes out of flat sheets, sort of like computer-controlled origami.

    These machines are expensive and are mostly targeted at big companies. But technology gets cheaper over time, and already machines are available in the $15-20,000 price range, the same kind of price that a colour laser printer would have been a decade ago.

    And rapid-prototyping is only a small step away from real manufacturing. In the software world, tools such as Visual Basic or REALbasic that were originally marketed as rapid prototyping aids are now used to develop final version of multi-million dollar software packages.

    The fundamental problem with rapid prototyping machines is mostly one of materials. A 2D colour printer can print any colour imaginable by mixing (or interleaving) a mere 4 colours of ink. But if you want to "print" a mechanical device, it's not sufficient to get the right colour, you have to use the right materials.

    A lamp might be made of porcelain, metal and fabric. A computer is made from plastics, various metals, silicon and half-a-dozen other materials. To further complicate matters, a computer contains components such as microchips that must be manufactured in a clean room with incredible precision - not really suitable candidates for desktop "printing".

    But there is hope. General purpose programmable analogue or digital microchips have been available for some time that can emulate any circuit or custom-manufactured chip's behaviour. These are mass-produced and can cost as little as a dollar. Techniques have recently been developed to print circuit boards cheaply and easily, and rapid prototyping machines and automated milling machines are constantly being improved to work with a greater variety of increasingly durable materials.

    A Star Trek replicator or cornucopia machine (from Charles Stross' Singularity Sky) is still a long way off, but it is becoming feasible that a multipurpose manufacturing machine could be built that has several specialised printing heads and multiple raw material imputs (one slot might take a hopper full of generic ICs, another might be a tube of resin or plastic beads, a third might be metal filings, and so on).

    This would be able to construct a wide range of possible consumer electronics devices to a standard close to that of normal manufacturing. And it wouldn't matter if the end product wasn't as durable as a "real" product, because if it broke down you could just print another - heck there'd probably be a new model available to download by then anyway!

    The acid test for a general-purpose programming language is whether it can be used to write its own compiler, and it seems that we are already nearing the achievement of the equivalent benchmark with 3D printing technology, namely a printer that can print itself.

    A printer that can print itself takes us neatly onto the final candidate technology of the bespoke hardware revolution...

  • Autonomous Replicators.

    So far merely the stuff of science fiction, these will one day be the last word in scalable manufacturing.

    Replicators, also know as Von Neuman machines (after the mathematician John Von Neuman who initially conceived the idea), are macro-scale or nano-scale robots that are able to reproduce themselves from raw materials. This is actually quite plausible in the short term if the "raw materials" in question are just specially designed generic components - essentially these devices would be like mobile 3D printers. But to fit the definition properly they've got to be able to reproduce themselves from scratch.

    And for that it doesn't matter whether the replicator itself is the size of a bus or smaller than a pinhead because it's going to have to be able to manipulate individual atoms, and that means nanotechnology.

    Nanotechnology is machinery that is built to a precision of individual atoms, and is capable of manipulating matter at the same scale. A replicator that employed nanotech would be able to build perfect copies of anything you could imagine, and at any scale. If it needed more workers it could just replicate itself. If it needed more materials it could convert its own colleagues back into raw materials. And irrespective of the size of the product, it would still be able to build with atomic precision - it could potentially build a skyscraper out of a single crystal of diamond.

    Replicators could potentially form both the factory and components of a machine. In Tony Ballantyne's Recursion, spider-like replicators are used to construct buildings by essentially holding hands and forming themselves into the necessary structure before being welded together.

    Don't hold your breath though - replicators rank right up their with artificial intelligence (AI) and nuclear fusion - all of them technologies which we know are possible because they already occur in nature, but we've no idea how to achieve them yet (it's no coincidence that the best sources on the subject are science fiction novels).

Conclusions

So there we have it. Two current technologies (and one rather speculative) that are teetering on the brink of mass-market viability - perfect environmental conditions for a technological revolution. All that is needed now is that mysterious impetus that makes a technology go from a working prototype to the must-have item for every household.

The ability to manufacture hardware, especially consumer electronics, cheaply in the home is going to happen soon, and it will totally change the relationship between consumers and service providers.

When the Internet happened, and distributed entertainment media went digital, it started changing everything we previously knew about concepts such as intellectual property. Suddenly it was possible to make perfect copies of any published work, essentially for free.

Software piracy is forcing distributors to look to new ways of monetising media. Meanwhile, the flaws in the software patents system are having a similar effect on software developers. Originally designed to promote innovation, software patents now seem to serve no purpose but to allow software monopolies to strong-arm their competitors, or make lawyers rich as big companies sue each other over issues such as who invented the double click. Because of this, people are beginning to question whether these laws have a place in post-Internet society.

Until now, hardware has been mostly free of these concerns. You can't "copy" an electronic device unless you have access to huge manufacturing facilities, and a research team to reverse engineer the technology. Some people may quibble that one MP3 player's interface is too similar to another's, but mostly nobody would risk an out-and-out hardware rip-off because there's too much money at stake if they get caught out.

I've stated before what a smart decision it was for Apple to go the route of focussing on hardware rather than relying on software revenue. As Microsoft struggles with piracy, and trying to convince the consumer that DRM is good for them, Steve Jobs can relax on a beach as the cash comes rolling in from iPod and MacBook sales.

This is going to change. In the not-too distant future, kids will be downloading iPod schematics off the Internet, installing DRM-removal patches and .ogg codecs, changing the colour to purple and then printing them out on their bedroom printer.

Even if companies somehow prevent their design schematics from being released then the potential for bootlegging brand names will be enormous; Can't afford an iPod? Just download an open-source MP3 player and draw an Apple logo on the casing before you print it - your friends will never know the difference!

Hardware designers won't be able to rest on their laurels with a product - only constant, breakneck innovation will allow commercial organisations to compete against the tide of freeware and open-source copycat products. Some companies won't be able to do this and will fold, but others will thrive and the consumer only stands to benefit in the end.

I don't pretend to know exactly when this will happen, it may be in five years, it may be another couple of decades, but it will happen, and the only question remaining is whether the industry will be ready for it when it does.

 

Disclaimer: The opinions expressed here are those of the author and are not shared by Charcoal Design unless specifically stated. The material is for general information only and does not constitute investment, tax, legal or other form of advice. You should not rely on this information to make (or refrain from making) any decisions. Always obtain independent, professional advice for your own particular situation.

Comments

3d printer revolution

One problem I see is this: computers are universal computers. But 3d printers aren't universal constructors and aren't going to be for a long time yet. Hence they will only be able to make crude objects.

Perhaps they can make another 3d printer. But can they make another normal printer? How will they make the ink?

How will they make a pair of shoes? They can make a perfect fit, but what about the material? Can a material comfortable for shoes be synthesised out of whatever standard materials are in its input bottles?

If they can only make crude objects, can they be a big revolution?

-- D

Reply

Posted by David at 1:36pm on 09 Mar 2007

Re: 3d printer revolution

It is true that no printer is going to be capable of universal construction for some time yet, however that doesn't prevent them becoming successful within a limited scope of application.

Many consumer products such as shoes and sofas have a relatively slow upgrade cycle and demonstrate relatively little reason for companies' intentions to conflict with those of consumers.

Other than artificially inflated prices (or moral outrage at sweatshops, etc), I can't see any reason why anyone would "rebel" against Nike or Reebok. And if price is an issue it's already possible to buy much cheaper trainers that don't come with a designer label.

I see these devices as being initially of more interest for manufacturing consumer electronics that aren't encumbered by DRM, or subject to security concerns such as viruses. They will be used to make open source MP3 players and games consoles that don't bother the user with irritations such as format and region restrictions, or put them at the whim of arbitrary decisions made by the manufacturers such as "300 song limit" or "no video capability on the low-end model".

We have already seen open firmware replacements being released for devices such as the iPod, or the Lego RCX unit. I see open hardware as the next logical step.

And with "printable" circuits and displays both becoming a reality, and the fact that the most complex parts of electronics devices are inherently constructed from 2D layers anyway, I think it is quite likely that we will see printers for these emerging soon.

I also think that the availability of CAD software for driving these devices is going to popularise a fashion for DIY designing of other goods, which can then be sent to mail-order machine shops for manufacture.

Reply

Posted by Nick at 2:25pm on 09 Mar 2007

Re: Re: 3d printer revolution

Nick wrote:

"And with "printable" circuits and displays both becoming a reality, and the fact that the most complex parts of electronics devices are inherently constructed from 2D layers anyway, I think it is quite likely that we will see printers for these emerging soon."

Hmm. Won't the professional ones still always be orders of magnitude better, until such time as Moore's Law fails? And therefore, won't they always be able to provide the cooler products -- at perhaps the same price or cheaper? (Albeit less customised.)

-- D

Reply

Posted by David at 3:12pm on 09 Mar 2007

Re: 3d printer revolution

David wrote:

"Hmm. Won't the professional ones still always be orders of magnitude better, until such time as Moore's Law fails? And therefore, won't they always be able to provide the cooler products -- at perhaps the same price or cheaper? (Albeit less customised.)"

Yes, well that's certainly true of 2D printers. I guess that's why nobody ever bought any... ;-)

Printing your own anything will probably never be both cheaper and higher quality than buying a mass produced item, and yet there are many successful DIY products available, such as typewriters, paintbrushes, bread makers, sandwich makers (or even ovens for that matter), inkjet printers, etc. For the most part, these devices produce inferior copies of commercial products, and at a higher per-unit cost. Even the humble television is basically just a poor man's DIY cinema. But this has in no way been a barrier to the success of these devices.

All of these are ways of achieving something at home that you could buy in a shop for less money, or at a higher quality, or even just for less effort (though rarely all three).

Initially, 3D printers won't be aiming to compete on price. I believe that the initial drive for them may be intellectual property rights issues. As the situation with patents and the DMCA gets worse, there is a very real possibility that it will become impractical (or even illegal) to sell all kinds of interesting consumer electronics devices because of their potential for being used to violate copyright, or because the patents covering parts of their design would make the profit margin impractical.

But these restrictions won't affect home-brew hardware, nor will they prevent people distributing blueprints for free via the internet.

There's also another issue. A 3D printer than can print upgrades to itself is quite unlike any device currently in existence. With ordinary hardware, you purchase it hoping that the company that built it will continue to provide support for it in the form of spare parts, upgrades and peripherals. If they don't you eventually have to give up on it. For a 3D printer, and products that can be created with it, this ceases to be a concern. You can print your own upgrades and replacement parts. If a component breaks down, you can print a new one on your neighbour's machine.

Reply

Posted by Nick at 3:55pm on 09 Mar 2007

Post a Comment

(optional)

Plain text only - html tags are not supported.