Shooey thanks for the update!

Any pictures of your knew last?

I wish I had the money  to buy a scanner and CNC machinery - that's my dream. With a mansion to put it all in

A remaining challenge is that the last is not an exact replica of the foot.  While the measurements (circumference/girth) may be the same as the foot, the shape of the last differs from that of the foot.  In some cases, even the measurements of the last differ from those of your foot.  For example, the top of the last at the back (the comb) is thinner than the corresponding part of your foot.  Of course, the last also has an additional length in front to provide space for your toes.  Likewise, the bottom surface of the last is not the same as the bottom of your foot.  So, the reality is that any 'system,' be it computer or human (lastmaker), must take the data from your foot and manipulate/modify it for the production of the last.

Further, remember that you are manipulating a 3 dimensional object (CAD/CAM file); as such, I don't think it is easy to write a computer program to automatically modify an existing 3D shape based on the measurements of your foot and how those compare to the existing 'model.'  I have seen some lasts produced by altering an existing last to incorporate new measurements (done not automatically, but through an experienced user manipulating the file using a CAD system), and the results were far from elegant.  Moreover, the cost of the paying the operator to manipulate the file was anything but trivial, as trained CAD/CAM operators ain't cheap.

There is also the additional issue of the nature of the foot (bony vs. fleshy, stiff vs. flexible) which can be identified by an experienced lastmaker and incorporated into the shape and measurements of the last.  Similarly, there is a judgment available to the lastmaker as to how much the toes can be 'clipped' when shaping the toe of the last.  Finally, there is the customer's preference, in terms of how tight a fit the customer prefers, which a lastmaker can assess when measuring the foot by tightening/loosening the tape and reflect in the finished last.  I don't think these issues are going to be incorporated into any computer measuring system, and certainly not one that features broadly distributed scanners operated by non-lastmakers.

Hence, the challenge is not getting the foot's data into the system, nor is it producing a last from a CAD/CAM file, it is in the translation and manipulation of the data taken from the foot into a desirable shape and measurements for the last.

Consequently, at least at present, there will remain the need for an experienced lastmaker to be involved in the production of a last, if you want a consistently bespoke-quality result, regardless of whether that last is done by hand or is created via a CAD/CAM file on a computer for subsequent milling.

To date, most computer measuring system are used for a 'best-fit' approach, where the foot is compared to a library of existing lasts, and a best fit is selected from the range of options.  I believe both JM Weston and Lobb Paris pursued a computer-driven scanning/measurement system for a bespoke line, but I think both have thrown in the towel on this.

Shooey, I hope you can fill us in further on how Bruce's system works, and what he is hoping to do with it.  Perhaps he has plans which address some of the challenges I cite -- I certainly would love to learn what he says about all of this.

Alot of really good points made here. 

A also want to stress how difficult it is to find experienced CNC operators. I once wanted to have very simple shapes cut in pieces of stone and to have this done locally would have cost a fortune and a half. It ended 5 times cheaper to have the stone shipped to some stone shop in china, worked by hand and send it back to me.

Automation is very important in industry when you want to make many copies, in which case it becomes cost effective. But make no mistake, lasts will never be produced in an automated way without an old school skilled last maker very close by.

It sounds like Bruce Millar offers a true solution as he IS a skilled old school last maker that uses the machines to speed up the process. For instance if he has a last that he finds perfect, he can have this scanned and then have several identical copies produced, something that is tedious to do by hand (they copies are never perfectly identical either). He can also have a less than perfect last produced and then fine tune this by hand afterwards.

Since my Makerbot Replicator did cost ca. 1500 € and pedographie is standard at least in orthopedic lastmaking I doubt the 100,000 $ fairytale. Of course, fumbling around with Gcode takes some practice, but since the general shape of a foot doesnt change I can see no major problems and doubt that a 3d laserscan is even neccesary.

Last edited by formby (2012-09-17 08:37:06)

Last edited by xenon1 (2012-09-17 08:54:23)

This entire topic is of great interest to me and is one I've spent a fair amount of time contemplating.

There is no doubt that milling machines, lathes, CAD/CAM, etc. have great value in many applications.  However, in general, the greatest value is in producing identical items in large numbers.  (This is not the case with makerbots.)

Further, there is no doubt that you can design a last using a CAD/CAM system, and likewise you can modify an existing last using these same systems.

However, when you start to address the question of using said technologies for custom lasts, I think the technology still has quite a way to go.  For this application, you need to address two issues: cost and quality.  While the Delcam Crispin lastmaker app cited earlier certainly looks interesting, I imagine the capital investment needed for this would be north of $50K.  Add in to that the required cost of training an operator, and then additional expense for actually figuring out how to get good results, and you would be looking at a substantial investment.  Given the small scale of most (all?) bespoke shoemakers, I think you can understand why these systems are not deployed in the bespoke shoemaking trade.  For a really well trained, experience lastmaker, it probably takes say 4 hours to make a pair of bespoke lasts (arguably, it can be more like 2, but let's be generous here).  If the Delcam system would take 1/2 hour of operator time to come up with a pair of bespoke lasts, that's a saving of 3.5 hours; assuming, arguendo, that the lastmaker and the CAD/CAM operator are paid the same, say $100k/year, then the savings are 3.5 hours x ($100,000/year / 2000 hrs/year) = $175/pair.  To amortize the $75,000 of investment (omitting any opportunity cost of money) would require production of 428 pairs of lasts, and even more if the lastmaker is efficient or if the CAD/CAM operator is paid more.  For most bespoke firms, that is a alot of pairs of lasts.

Then, there is the quality issue; I remain dubious about the quality that these systems produce in one-off, custom last production.  Springline, which is mentioned earlier, is certainly a quality operation, and they know their lastmaking.  They have very modern, computer-driven lathes and CAD/CAM systems for their production work -- making lasts for the various UK factories, making shoe trees for the same, and making rough- and fine-turn blanks for the bespoke trade.  However, if you go to them for a pair of bespoke lasts, they have a couple of lastmakers who still take your measurements and tracings by hand, and they make your lasts by hand.  I am sure they have researched the alternatives, but to date they still do it the old fashioned way.

Then, there remains the challenge of 'data capture' from the feet.  Traditional lastmakers prefer to take your measurements and inspect your feet themselves; they then get feedback based on fitters and finished shoes as to how well their measurement system and production approach actually work for customers' feet.  The decoupling of the data capture from the finished product (the completed shoe, not just a last) is something these systems need to contemplate; otherwise the feedback loop necessary for continuous improvement of the lastmaking process is not going to be very satisfactory.

That is not to say the world won't head down this path; computer technology improves and innovation continues, and the costs keep coming down.  At the same time, the opportunity for people to learn custom lastmaking is very limited, so it may be we end up with the computerized solutions by default.  I would certainly love to be able to experiment with the Delcam Lastmaker system; I'm sure it would be fun.  I would like to see if one can design on it the way the bespoke lastmakers do things by hand -- from looking at the Delcam videos, the steps they show in modifying the lasts are not the same as how lastmakers go about making bespoke lasts.  However, to date the market seems to indicate the traditional approach has a better quality/cost tradeoff.

The problem with makerbots (putting aside the challenge of generating the appropriate design of a last) is the cost of the materials.  A pair of lasts uses a lot of plastic, probably several hundred dollars worth.