Sunday, 20 October 2013

Economist: 3D Printing Scales Up

I’m excited about 3D printing technology and believe it will have an important role to play in manufacturing.  The Economist recently had a good article on the current state of the technology.  Below is a highlight of the article and some of my thoughts.


Traditional manufacturing, the kind generally seen in factories today, operates thus: take a piece of material, chip/gouge away the unwanted pieces and/or bend/mold it until a finished component emerges, similar to a sculptor chipping away at a piece of rock to reveal a sculpture.  Various components are then assembled if necessary.  This “chipping away” gives it the name “subtractive” manufacturing.

3D printing, or “additive” manufacturing, takes the opposite approach.  As a construction worker builds a house by laying down a brick on top of another, using cement as “glue”.  3D printers builds the final 3D object by printing very thin layers of cross sections of the product on top of each other, bond together by an adhesive.  The analogy isn’t perfect, but hopefully the point is clear.

Important Characteristics of 3D Printing

The basic said difference entail the following advantage of 3D printing:

  • Expanded shape possibilities.  Certain complex shapes are very difficult or virtually impossible to make by chipping away.  When possible, such a part often needed to be broken into much smaller parts, made separately, then assembled.  Since 3D printing works by printing the cross sections of objects and aren’t constrained in the same way, designers can focus more on producing shapes or structures that optimize the purpose of the product.  For example, produce paddings with a certain lattice structure that absorbs impact much better, or jet engine parts that allow for higher efficiency by employing complex internal shapes.  Or, perhaps in the future, print organs to be implanted in patients (which are too complex in shapes to produce now) using bio-compatible materials
  • One-piece construction.  Whereas subtractive manufacturing often needs to make various components independently and then assembled, additive manufacturing can make the entire product as an integrated whole.  This one-piece construction have the potential to make the final products stronger and more durable.
  • Reduced waste.  A lot of “chipped away” material can result from traditional manufacturing, whereas the additive process only lays down what will be the finished product, so less is wasted – in some cases, much less.

Furthermore, 3D printing has the following advantage:

  • Lower scale requirement.  Traditional manufacturing requires scale for the products to be cost accessible.  To produce any finished components, manufacturing equipment often have to be retooled and production line reconfigured to put all the pieces together.  This one-time set up cost is often too prohibitive for producing small batches and takes a long time.  3D printing, just as a 2D inkjet printer can print any shape on paper, doesn’t require this set up, and so does not require the same scale for a feasible price on the finished products.  It can start producing products in a timely manner.

What’s interesting to me is both manufacturing methods have their merits and can complement each other.  Traditional manufacturing, once set up properly, can make products at high speeds and low costs that 3D printing can’t come close to matching.  But for small scale, highly specialized and high value added parts, the removal of the set up cost in time and money becomes very beneficial.  Out of production parts can thus be made in days rather than weeks, for example.  3D printing therefore satisfies an unmet demand in the marketplace.

I also see that in a sense, 3D printing is reducing the “hardware problem” of retooling the machines and reconfiguring the production line to more of a “software problem” – as long as a product can be designed in a software program that can be understood by a 3D printer, it can be made.  It reduces the required knowledge for the manufacturing process.


3D printing’s ability to produce complex shapes, reduce waste, fulfill the demand for small scale production and complement the traditional manufacturing process excites me.  Though it is still limited by the number of materials that are 3D printable, and its speed is still slow, they will improve with time, and I look forward to its wider application.

Friday, 24 May 2013

Video: 3 part discussion about the J S Bach's Matthews Passion

With Ton Koopman, Christoph Wolff, Petrer Wollny, Martin Petzold.

Bill Gate’s science and innovation top reads

Bill Gates recently shared his recommended books in science and innovation in his blog post.  Among them are Where Good Ideas Come From by Steven Johnson and Energy Myths and Realities by Vaclav Smil.

I will be adding a link to this post in my reading list.

Monday, 13 May 2013

The Economist: Driverless Cars – Look, no hands

Driverless cars is a very interesting field to me, and this week The Economist posted a good article on the prospects of this technology.

To me, here are a few things about electronic-fy the car that excites me:

  1. Sensors/Analytics: the car can now serve as an endpoint for collecting ever precise data about traffic and driving patterns for analysis and optimization, be it public urban planning or private insurance offerings
  2. Autonomy: with the cars able to drive themselves, automobiles are no longer bounded by the movement of its owners the way they are today, and with the ability to never forget learned lessons, algorithms will get better and better at driving over the long run also
  3. Interconnectedness: the ability to communicate not just with centralized points such as emergency services but with other cars in proximity allows the car “network” to be more responsive in human mobility needs

In all, the potential for advances in such fields to turn automotive from an ownership to a service paradigm/business model is quite exciting.  Driving is getting more difficult, and as has happened before, once rules for certain aspect of driving becomes concrete enough and algorithms can outperform a human driver in decision quality, while doing so with a feasible price tag, the technology will make more inroads to popular use.

When we get to that point, perhaps a few more interesting questions will then arise.  Could it be possible that the hardware of a common sedan be so commoditized that a large portion of the value of a car come from the software?  Long time ago, IBM thought hardware constituted the core value of the computing industry.  Could the same be said of cars now or in the future?  What role will software play in the long run on the dynamic and diverse behaviour of the future cars to come?

Interesting questions indeed.