Industrial Session

In the industrial session, speakers will share their expertise in industries related to various applications of geometric modelling and processing, and in particular focus on challenges and unsolved problems within their industrial domains.

Adrian Bowyer (RepRap Pro)

Ever since we first hit one flint with another, tools have been used to make tools.

Interestingly, the tools made can be more accurate than the tools used to make them as long as the made tools are adjustable and we have a means of measuring their output accurately. Since the Industrial Revolution machine tools have had screw adjusters to get them square and to set critical internal distances.

It is quite easy to make a tool like a 3D printer at very low cost that is both stable and that moves repeatably. This is made even easier by the fact that 3D printing requires very low forces (unlike metal cutting, for example). But making a low-cost 3D printer linear and square is much harder. Like most modern tools, 3D printers are driven by a computer, so - instead of screw adjusters - we can correct for geometric errors by mapping the tool movements requested by the user into the space of the tool so that 3D printed objects come out right.

This talk will look at some simple geometric corrections like this that have already been implemented, and then discuss the problems that remain to be solved.

Short Bio:

Adrian Bowyer holds a first degree and a PhD in engineering from Imperial College. He was an academic engineer and mathematician at the University of Bath for 35 years, from where he retired in 2012 to become a director of RepRap Professional Ltd.

His areas of research are geometric modelling and geometric computing in general (he is one of the authors of the Bowyer-Watson algorithm for Voronoi diagrams), the application of computers to manufacturing, and biomimetics. He is the author of about one hundred papers and books on many different aspects of engineering, computing, mathematics and biology. In 2004 he created RepRap - humanity's first self-replicating general-purpose manufacturing machine.

Jan Vandenbrande (Boeing)

I will cover an assortment of challenges that we are facing in aerospace geometric design and manufacturing:

• The problem with solid modeling is that parts are not "solid" (i.e., not infinitely rigid)
• Modeling composites is hard (our version of "big data")
• It all starts with materials: How can we tailor material properties to suit our needs? How do we leverage new manufacturing technologies such as additive manufacturing?
• Geometry is math, we should treat is as such.

Short Bio:

WJan Vandenbrande is a Technical Fellow and Senior Manager of Boeing's Applied Math Geometry and Optimization group. Jan leads a team to create advanced geometry and optimization tools to conduct design trade studies on air and space vehicles, and to solve difficult geometric processing problems for composite layups. Jan has also created several systems to analyze and automate machining based on his academic work and developed several parametric aircraft for design studies. This body of work is having a major business impact.

Before joining Boeing, Jan was a select member of a team to define the next generation CAM system and to enhance system performance at Unigraphics, now called Siemens NX. Jan obtained a PhD in electrical engineering from the University of Rochester for his work on machinable feature recognition completed at the computer science department of the University of Southern California under Prof Requicha. Jan is a past Associate Editor of 3 professional journals; served on the executive board of the Solid Modeling Association; published over half a dozen papers; invited speaker at several conferences; and holder of several patents.