TEDSalon London Spring 2011
Lisa Harouni: A primer on 3D printing
Filmed:
Readability: 4.7
1,788,111 views
2012 may be the year of 3D printing, when this three-decade-old technology finally becomes accessible and even commonplace. Lisa Harouni gives a useful introduction to this fascinating way of making things -- including intricate objects once impossible to create.
Lisa Harouni - 3D printing entrepreneur
Lisa Harouni is the co-founder of Digital Forming, working in "additive manufacturing" -- or 3D printing. Full bio
Lisa Harouni is the co-founder of Digital Forming, working in "additive manufacturing" -- or 3D printing. Full bio
Double-click the English transcript below to play the video.
00:15
It is actually a reality today
0
0
3000
00:18
that you can download products from the Web --
1
3000
4000
00:22
product data, I should say, from the Web --
2
7000
3000
00:25
perhaps tweak it and personalize it
3
10000
3000
00:28
to your own preference or your own taste,
4
13000
2000
00:30
and have that information sent
5
15000
3000
00:33
to a desktop machine
6
18000
3000
00:36
that will fabricate it for you on the spot.
7
21000
2000
00:38
We can actually build for you,
8
23000
2000
00:40
very rapidly,
9
25000
2000
00:42
a physical object.
10
27000
3000
00:45
And the reason we can do this
11
30000
2000
00:47
is through an emerging technology
12
32000
2000
00:49
called additive manufacturing,
13
34000
2000
00:51
or 3D printing.
14
36000
3000
00:54
This is a 3D printer.
15
39000
2000
00:56
They have been around
16
41000
2000
00:58
for almost 30 years now,
17
43000
2000
01:00
which is quite amazing to think of,
18
45000
2000
01:02
but they're only just starting
19
47000
2000
01:04
to filter into the public arena.
20
49000
2000
01:06
And typically, you would take data,
21
51000
3000
01:09
like the data of a pen here,
22
54000
3000
01:12
which would be a geometric representation of that product in 3D,
23
57000
3000
01:15
and we would pass that data with material
24
60000
2000
01:17
into a machine.
25
62000
2000
01:19
And a process that would happen in the machine
26
64000
2000
01:21
would mean layer by layer that product would be built.
27
66000
2000
01:23
And we can take out the physical product,
28
68000
3000
01:26
and ready to use,
29
71000
2000
01:28
or to, perhaps, assemble into something else.
30
73000
3000
01:31
But if these machines have been around for almost 30 years,
31
76000
3000
01:34
why don't we know about them?
32
79000
2000
01:36
Because typically they've been too inefficient,
33
81000
3000
01:39
inaccessible,
34
84000
3000
01:42
they've not been fast enough,
35
87000
2000
01:44
they've been quite expensive.
36
89000
2000
01:46
But today,
37
91000
2000
01:48
it is becoming a reality
38
93000
2000
01:50
that they are now becoming successful.
39
95000
2000
01:52
Many barriers are breaking down.
40
97000
2000
01:54
That means that you guys
41
99000
2000
01:56
will soon be able to access one of these machines,
42
101000
2000
01:58
if not this minute.
43
103000
3000
02:01
And it will change and disrupt
44
106000
2000
02:03
the landscape of manufacturing,
45
108000
2000
02:05
and most certainly our lives, our businesses
46
110000
3000
02:08
and the lives of our children.
47
113000
2000
02:10
So how does it work?
48
115000
3000
02:13
It typically reads CAD data,
49
118000
2000
02:15
which is a product design data
50
120000
2000
02:17
created on professional product design programs.
51
122000
2000
02:19
And here you can see an engineer --
52
124000
2000
02:21
it could be an architect or it could be a professional product designer --
53
126000
3000
02:24
create a product in 3D.
54
129000
3000
02:27
And this data gets sent to a machine
55
132000
2000
02:29
that slices the data
56
134000
3000
02:32
into two-dimensional representations of that product
57
137000
2000
02:34
all the way through --
58
139000
2000
02:36
almost like slicing it like salami.
59
141000
3000
02:39
And that data, layer by layer, gets passed through the machine,
60
144000
3000
02:42
starting at the base of the product
61
147000
2000
02:44
and depositing material, layer upon layer,
62
149000
3000
02:47
infusing the new layer of materials to the old layer
63
152000
4000
02:51
in an additive process.
64
156000
2000
02:53
And this material that's deposited
65
158000
3000
02:56
either starts as a liquid form
66
161000
2000
02:58
or a material powder form.
67
163000
3000
03:01
And the bonding process can happen
68
166000
2000
03:03
by either melting and depositing or depositing then melting.
69
168000
3000
03:06
In this case, we can see a laser sintering machine developed by EOS.
70
171000
3000
03:09
It's actually using a laser
71
174000
2000
03:11
to fuse the new layer of material to the old layer.
72
176000
3000
03:14
And over time --
73
179000
2000
03:16
quite rapidly actually, in a number of hours --
74
181000
2000
03:18
we can build a physical product,
75
183000
2000
03:20
ready to take out of the machine and use.
76
185000
2000
03:22
And this is quite an extraordinary idea,
77
187000
4000
03:26
but it is reality today.
78
191000
3000
03:29
So all these products that you can see on the screen
79
194000
3000
03:32
were made in the same way.
80
197000
2000
03:34
They were all 3D printed.
81
199000
2000
03:36
And you can see,
82
201000
2000
03:38
they're ranging from shoes,
83
203000
2000
03:40
rings that were made out of stainless steal,
84
205000
3000
03:43
phone covers out of plastic,
85
208000
2000
03:45
all the way through to spinal implants, for example,
86
210000
3000
03:48
that were created out of medical-grade titanium,
87
213000
2000
03:50
and engine parts.
88
215000
2000
03:52
But what you'll notice about all of these products
89
217000
2000
03:54
is they're very, very intricate.
90
219000
2000
03:56
The design is quite extraordinary.
91
221000
3000
03:59
Because we're taking this data in 3D form,
92
224000
3000
04:02
slicing it up before it gets past the machine,
93
227000
3000
04:05
we can actually create structures
94
230000
2000
04:07
that are more intricate
95
232000
2000
04:09
than any other manufacturing technology --
96
234000
3000
04:12
or, in fact, are impossible to build in any other way.
97
237000
4000
04:16
And you can create parts with moving components,
98
241000
3000
04:19
hinges, parts within parts.
99
244000
2000
04:21
So in some cases, we can abolish totally
100
246000
3000
04:24
the need for manual labor.
101
249000
2000
04:26
It sounds great.
102
251000
2000
04:28
It is great.
103
253000
2000
04:30
We can have 3D printers today
104
255000
2000
04:32
that build structures like these.
105
257000
2000
04:34
This is almost three meters high.
106
259000
3000
04:37
And this was built
107
262000
2000
04:39
by depositing artificial sandstone layer upon layer
108
264000
3000
04:42
in layers of about five millimeters to 10 mm in thickness --
109
267000
4000
04:46
slowly growing this structure.
110
271000
2000
04:48
This was created by an architectural firm called Shiro.
111
273000
3000
04:51
And you can actually walk into it.
112
276000
2000
04:53
And on the other end of the spectrum,
113
278000
2000
04:55
this is a microstructure.
114
280000
2000
04:57
It's created depositing layers
115
282000
2000
04:59
of about four microns.
116
284000
3000
05:02
So really the resolution is quite incredible.
117
287000
3000
05:05
The detail that you can get today
118
290000
2000
05:07
is quite amazing.
119
292000
3000
05:10
So who's using it?
120
295000
2000
05:12
Typically, because we can create products very rapidly,
121
297000
3000
05:15
it's been used by product designers,
122
300000
3000
05:18
or anyone who wanted to prototype a product
123
303000
3000
05:21
and very quickly create or reiterate a design.
124
306000
3000
05:24
And actually what's quite amazing about this technology as well
125
309000
4000
05:28
is that you can create bespoke products en masse.
126
313000
3000
05:31
There's very little economies of scale.
127
316000
3000
05:34
So you can now create one-offs very easily.
128
319000
3000
05:37
Architects, for example,
129
322000
2000
05:39
they want to create prototypes of buildings.
130
324000
2000
05:41
Again you can see,
131
326000
2000
05:43
this is a building of the Free University in Berlin
132
328000
3000
05:46
and it was designed by Foster and Partners.
133
331000
2000
05:48
Again, not buildable in any other way.
134
333000
3000
05:51
And very hard to even create this by hand.
135
336000
3000
05:56
Now this is an engine component.
136
341000
3000
05:59
It was developed by a company called Within Technologies
137
344000
3000
06:02
and 3T RPD.
138
347000
3000
06:05
It's very, very, very detailed
139
350000
2000
06:07
inside with the design.
140
352000
2000
06:09
Now 3D printing
141
354000
2000
06:11
can break away barriers in design
142
356000
2000
06:13
which challenge the constraints
143
358000
2000
06:15
of mass production.
144
360000
2000
06:17
If we slice into this product which is actually sitting here,
145
362000
3000
06:20
you can see that it has a number of cooling channels pass through it,
146
365000
4000
06:24
which means it's a more efficient product.
147
369000
3000
06:27
You can't create this with standard manufacturing techniques
148
372000
3000
06:30
even if you tried to do it manually.
149
375000
2000
06:32
It's more efficient
150
377000
2000
06:34
because we can now create all these cavities within the object
151
379000
3000
06:37
that cool fluid.
152
382000
2000
06:39
And it's used by aerospace
153
384000
2000
06:41
and automotive.
154
386000
3000
06:44
It's a lighter part
155
389000
3000
06:47
and it uses less material waste.
156
392000
2000
06:49
So it's overall performance and efficiency
157
394000
3000
06:52
just exceeds standard mass produced products.
158
397000
3000
06:55
And then taking this idea
159
400000
2000
06:57
of creating a very detailed structure,
160
402000
2000
06:59
we can apply it to honeycomb structures
161
404000
2000
07:01
and use them within implants.
162
406000
3000
07:04
Typically an implant
163
409000
2000
07:06
is more effective within the body
164
411000
2000
07:08
if it's more porous,
165
413000
2000
07:10
because our body tissue will grow into it.
166
415000
2000
07:12
There's a lower chance of rejection.
167
417000
3000
07:15
But it's very hard to create that in standard ways.
168
420000
3000
07:18
With 3D printing,
169
423000
2000
07:20
we're seeing today
170
425000
2000
07:22
that we can create much better implants.
171
427000
2000
07:24
And in fact, because we can create
172
429000
2000
07:26
bespoke products en masse, one-offs,
173
431000
3000
07:29
we can create implants
174
434000
2000
07:31
that are specific to individuals.
175
436000
2000
07:33
So as you can see,
176
438000
2000
07:35
this technology and the quality of what comes out of the machines is fantastic.
177
440000
4000
07:39
And we're starting to see it being used
178
444000
2000
07:41
for final end products.
179
446000
2000
07:43
And in fact, as the detail is improving,
180
448000
2000
07:45
the quality is improving,
181
450000
2000
07:47
the price of the machines are falling
182
452000
3000
07:50
and they're becoming quicker.
183
455000
2000
07:52
They're also now small enough
184
457000
2000
07:54
to sit on a desktop.
185
459000
2000
07:56
You can buy a machine today for about $300
186
461000
2000
07:58
that you can create yourself,
187
463000
2000
08:00
which is quite incredible.
188
465000
2000
08:02
But then it begs the question,
189
467000
2000
08:04
why don't we all have one in our home?
190
469000
3000
08:07
Because, simply, most of us here today
191
472000
2000
08:09
don't know how to create the data
192
474000
2000
08:11
that a 3D printer reads.
193
476000
2000
08:13
If I gave you a 3D printer,
194
478000
2000
08:15
you wouldn't know how to direct it
195
480000
2000
08:17
to make what you want it to.
196
482000
2000
08:19
But there are more and more
197
484000
2000
08:21
technologies, software and processes today
198
486000
2000
08:23
that are breaking down those barriers.
199
488000
2000
08:25
I believe we're at a tipping point
200
490000
2000
08:27
where this is now something
201
492000
3000
08:30
that we can't avoid.
202
495000
2000
08:32
This technology
203
497000
2000
08:34
is really going to disrupt
204
499000
2000
08:36
the landscape of manufacturing
205
501000
2000
08:38
and, I believe, cause a revolution
206
503000
2000
08:40
in manufacturing.
207
505000
2000
08:42
So today,
208
507000
2000
08:44
you can download products from the Web --
209
509000
2000
08:46
anything you would have on your desktop,
210
511000
2000
08:48
like pens, whistles, lemon squeezers.
211
513000
3000
08:51
You can use software like Google SketchUp
212
516000
3000
08:54
to create products from scratch
213
519000
2000
08:56
very easily.
214
521000
3000
08:59
3D printing can be also used
215
524000
2000
09:01
to download spare parts from the Web.
216
526000
3000
09:04
So imagine you have, say,
217
529000
2000
09:06
a Hoover in your home
218
531000
2000
09:08
and it has broken down. You need a spare part,
219
533000
3000
09:11
but you realize that Hoover's been discontinued.
220
536000
3000
09:14
Can you imagine going online --
221
539000
2000
09:16
this is a reality --
222
541000
2000
09:18
and finding that spare part
223
543000
2000
09:20
from a database of geometries
224
545000
2000
09:22
of that discontinued product
225
547000
2000
09:24
and downloading that information, that data,
226
549000
3000
09:27
and having the product made for you at home,
227
552000
2000
09:29
ready to use, on your demand?
228
554000
3000
09:32
And in fact, because we can create spare parts
229
557000
2000
09:34
with things the machines
230
559000
2000
09:36
are quite literally making themselves.
231
561000
2000
09:38
You're having machines fabricate themselves.
232
563000
3000
09:41
These are parts of a RepRap machine,
233
566000
2000
09:43
which is a kind of desktop printer.
234
568000
3000
09:46
But what interests my company the most
235
571000
3000
09:49
is the fact that you can create
236
574000
2000
09:51
individual unique products en masse.
237
576000
3000
09:54
There's no need to do a run
238
579000
2000
09:56
of thousands of millions
239
581000
2000
09:58
or send that product to be injection molded in China.
240
583000
3000
10:01
You can just make it physically on the spot.
241
586000
4000
10:05
Which means
242
590000
2000
10:07
that we can now present to the public
243
592000
2000
10:09
the next generation of customization.
244
594000
3000
10:12
This is something that is now possible today,
245
597000
2000
10:14
that you can direct personally
246
599000
2000
10:16
how you want your products to look.
247
601000
3000
10:19
We're all familiar with the idea
248
604000
2000
10:21
of customization or personalization.
249
606000
2000
10:23
Brands like Nike are doing it.
250
608000
2000
10:25
It's all over the Web.
251
610000
2000
10:27
In fact, every major household name
252
612000
2000
10:29
is allowing you
253
614000
2000
10:31
to interact with their products
254
616000
2000
10:33
on a daily basis --
255
618000
2000
10:35
all the way from Smart Cars
256
620000
2000
10:37
to Prada
257
622000
2000
10:39
to Ray Ban, for example.
258
624000
2000
10:41
But this is not really mass customization;
259
626000
2000
10:43
it's known as variant production,
260
628000
2000
10:45
variations of the same product.
261
630000
3000
10:48
What you could do is really influence your product now
262
633000
3000
10:51
and shape-manipulate your product.
263
636000
3000
10:54
I'm not sure about you guys,
264
639000
2000
10:56
but I've had experiences
265
641000
2000
10:58
when I've walked into a store and I've know exactly what I've wanted
266
643000
2000
11:00
and I've searched everywhere for that perfect lamp
267
645000
3000
11:03
that I know where I want to sit in my house
268
648000
2000
11:05
and I just can't find the right thing,
269
650000
2000
11:07
or that perfect piece of jewelry
270
652000
3000
11:10
as a gift or for myself.
271
655000
2000
11:12
Imagine that you can now
272
657000
2000
11:14
engage with a brand
273
659000
3000
11:17
and interact,
274
662000
2000
11:19
so that you can pass your personal attributes
275
664000
3000
11:22
to the products that you're about to buy.
276
667000
4000
11:26
You can today
277
671000
2000
11:28
download a product with software like this,
278
673000
2000
11:30
view the product in 3D.
279
675000
2000
11:32
This is the sort of 3D data
280
677000
2000
11:34
that a machine will read.
281
679000
2000
11:36
This is a lamp.
282
681000
2000
11:38
And you can start iterating the design.
283
683000
2000
11:40
You can direct what color that product will be,
284
685000
2000
11:42
perhaps what material.
285
687000
2000
11:44
And also, you can engage in shape manipulation of that product,
286
689000
3000
11:47
but within boundaries that are safe.
287
692000
2000
11:49
Because obviously the public are not professional product designers.
288
694000
3000
11:52
The piece of software will keep an individual
289
697000
3000
11:55
within the bounds of the possible.
290
700000
4000
11:59
And when somebody is ready to purchase the product
291
704000
2000
12:01
in their personalized design,
292
706000
2000
12:03
they click "Enter" and this data gets converted
293
708000
3000
12:06
into the data that a 3D printer reads
294
711000
4000
12:10
and gets passed to a 3D printer,
295
715000
5000
12:15
perhaps on someone's desktop.
296
720000
3000
12:18
But I don't think that that's immediate.
297
723000
2000
12:20
I don't think that will happen soon.
298
725000
2000
12:22
What's more likely, and we're seeing it today,
299
727000
2000
12:24
is that data gets sent
300
729000
2000
12:26
to a local manufacturing center.
301
731000
2000
12:28
This means lower carbon footprint.
302
733000
3000
12:31
We're now, instead of shipping a product across the world,
303
736000
3000
12:34
we're sending data across the Internet.
304
739000
2000
12:36
Here's the product being built.
305
741000
3000
12:39
You can see, this came out of the machine in one piece
306
744000
2000
12:41
and the electronics were inserted later.
307
746000
2000
12:43
It's this lamp, as you can see here.
308
748000
2000
12:45
So as long as you have the data,
309
750000
2000
12:47
you can create the part on demand.
310
752000
2000
12:49
And you don't necessarily need to use this
311
754000
2000
12:51
for just aesthetic customization,
312
756000
2000
12:53
you can use it for functional customization,
313
758000
2000
12:55
scanning parts of the body
314
760000
2000
12:57
and creating things that are made to fit.
315
762000
2000
12:59
So we can run this through to something like prosthetics,
316
764000
3000
13:02
which is highly specialized to an individual's handicap.
317
767000
3000
13:05
Or we can create very specific prosthetics
318
770000
4000
13:09
for that individual.
319
774000
2000
13:11
Scanning teeth today,
320
776000
2000
13:13
you can have your teeth scanned
321
778000
2000
13:15
and dental coatings made in this way to fit you.
322
780000
2000
13:17
While you wait at the dentist,
323
782000
2000
13:19
a machine will quietly be creating this for you
324
784000
3000
13:22
ready to insert in the teeth.
325
787000
2000
13:24
And the idea of now creating implants,
326
789000
4000
13:28
scanning data, an MRI scan of somebody
327
793000
3000
13:31
can now be converted into 3D data
328
796000
2000
13:33
and we can create very specific implants for them.
329
798000
4000
13:37
And applying this
330
802000
2000
13:39
to the idea of building up what's in our bodies.
331
804000
2000
13:41
You know, this is pair of lungs and the bronchial tree.
332
806000
3000
13:44
It's very intricate.
333
809000
2000
13:46
You couldn't really create this or simulate it in any other way.
334
811000
2000
13:48
But with MRI data,
335
813000
2000
13:50
we can just build the product,
336
815000
2000
13:52
as you can see, very intricately.
337
817000
2000
13:56
Using this process,
338
821000
2000
13:58
pioneers in the industry are layering up cells today.
339
823000
3000
14:01
So one of the pioneers, for example, is Dr. Anthony Atala,
340
826000
3000
14:04
and he has been working
341
829000
2000
14:06
on layering cells to create body parts --
342
831000
2000
14:08
bladders, valves, kidneys.
343
833000
4000
14:12
Now this is not something that's ready for the public,
344
837000
2000
14:14
but it is in working progress.
345
839000
3000
14:17
So just to finalize, we're all individual.
346
842000
2000
14:19
We all have different preferences, different needs.
347
844000
3000
14:22
We like different things.
348
847000
2000
14:24
We're all different sizes and our companies the same.
349
849000
2000
14:26
Businesses want different things.
350
851000
2000
14:28
Without a doubt in my mind,
351
853000
2000
14:30
I believe that this technology
352
855000
2000
14:32
is going to cause a manufacturing revolution
353
857000
2000
14:34
and will change the landscape of manufacturing as we know it.
354
859000
3000
14:37
Thank you.
355
862000
2000
14:39
(Applause)
356
864000
4000
ABOUT THE SPEAKER
Lisa Harouni - 3D printing entrepreneurLisa Harouni is the co-founder of Digital Forming, working in "additive manufacturing" -- or 3D printing.
Why you should listen
Lisa Harouni is the co-founder and CEO of Digital Forming, a company that works on the software side of 3D printing -- the design tools needed to run the new generaion of 3D printing processes. She has a background in economics, and worked in the G7 Economics team at Deutsche Bank AG before moving over to the consumer products business.
Lisa Harouni | Speaker | TED.com