I’ve praised FedEx before, but they still managed to overdo themselves. My contact person and the Finland country manager paid visit to me. I joked about the lazy and incompetent house lord (me) that had not dome the snow work for visitors early in the morning. When the guys left, they plowed the snow from my walkway and stairs to the street!
I’m working on the store right now. Please check back after a few hours.
I haven’t written much lately, shame on me! Anyway, this has been asked several times, so I’ll write the answer to here as well. So, what are my near term plans for the software? Bug fixes will get priority over new features, but here is my new stuff todo list, before feedback from this post, in the order of importance:
I am adding functions to the software, so I didn’t get to do a stable public release during the weekend, as I hoped. Once the new functions are stable (or at least don’t interfere with the current functionality so that the interface to those can be hidden without ill effects), the very next release will have support to 32mm tapes as well as yet another attempt to have a robust camera startup routine that is stable on all Windows and .NET versions.
Then, support for tape definitions not on the built in list, custom trays, feeders, trays and part jigs. I anticipate that IC placement will greatly improve when the pickup location can be accurately defined.
Placed / not yet placed information to jobs, as well as option to continue on error and making a list of failed components, if any.
Measure everything at the start of a job. This way, you can stay at the machine until it has all the the info a job needs and then walk away knowing, that a job will not fail in the middle to a failed measurement.
The “everything visible” style UI was clear (imo) when the software was released but is getting more and more cluttered when new functions are added. It needs a rearrangement, most likely resulting to a traditional menu bar and setup screens.
Some folks have machines with the capability of automated nozzle change, with different implementations. (I don’t, yet.) These should somehow be supported by the software. This has its own discussion thread here.
Due to spammers, commenting on my blog is disabled. (Do those folks really believe that my blog is widely read and they can get any audience between the time they post their spam and I get to remove the posts? Well, I guess the bots don’t care…) However, you can add your views either by registering on my forum and participating to this thread or by sending me email.
This blog is about the nature of Chinese manufacturing and selling business and how it appears and affects to small players outside China.
I haven’t travelled to China to arrange buying materials and parts for the machine; this is not big enough business to justify that. On some items, I have to go through dealers, which can make life difficult. The Ebay and Alibaba stores seem to come and go, and even the more stable dealers don’t get exactly the same item repeatedly. Running a small business myself, I understand them very well: If the item has the same description and specifications, there is little point making sure it is exactly the same – but the devil is in the details.
On some items I get or must to deal with the factories directly. I found out this the hard way: For example, if you have followed the story closely enough, you would have noticed that the type of chain kit has changed. I bought lots of different type of chains to find the best choice. After I decided which chain to use, I ordered plate set and chains for the first lot. When I got the chains, they were slightly different than my samples. That mishap was correctable by modifying the attachment piece. Although usable, the result was not at the quality I was aiming for. I didn’t want to scrap usable chains, but I didn’t want to sell them either. I ended up sending free chains with the first machine.
Next time I wasn’t so lucky. I redesigned the plate set to match the chains. The chains I got that time did not fit, no matter what. They did sort of fit the description in the on-line store, however. The local hackerspace got more chains than their robot and 3D printer projects needs and a metal recycling place got 20kg of steel. Bummer.
I can’t re-design and re-manufacture the laser cut plate set again every time I buy more chains to the store! So, I had to locate a factory that wants to deal with me directly, so I can keep getting exactly the item again. Fortunately I had the money and the storage room for a year (or more) cables for my store.
Lesson learned: It seems to be rather difficult to buy the same exact item again from China, especially when the item is considered to be a commodity. And in China, items that here would get a type number and a stable specification sheet can still be commodity items.
At the time of writing this, I’ve bought three lots of cameras. The store I first used is no more, and even though the description and appearance of the cameras in another store are exactly the same, the products are not. The USB identification string of the cameras changed a little. This did not make a difference in my application, but it is easy to imagine an application where it would.
What made a difference is that the “10mm outer diameter” specification is rather loose and different enough. Today, I had to add a note to instructions to use a piece of tape if the cameras don’t fit tightly to the holders. I’m not too happy about the situation, but I hope my customers agree the above to be a reasonable workaround in this case.
On the other hand, I knew this: I have been in high end audio business, and I know all too well that not all “10nF, size 1206, NP0” capacitors or even 10k resistors are the same, when you apply parts so that the unspecified characteristics (like above, the tolerance of 10mm) actually do matter.
I’m shipping my stuff using FedEx. My experiences are therefore mostly with FedEx Finland and whatever is happening in FedEx international offices, gets filtered by their Finland office. So, I can comment only from FedEx Finland office actions, but they do deserve a comment.
As I see it, this has to come from the top. Obviously, FedEx Finland management is doing something very right and maybe, hopefully, the origin is at the very top of FedEx management. So, here is my statement:
FedEx Finland has the best customer service I know of, and this is true no matter who is taking care of me. The driver I meet regularly is a happy person and never seems to have a bad day. Whenever I call or send email to FedEx, I always get my point through, and never, ever get a response “I can’t do anything” or “this is how the system works”. Even when an error is on my side and it is me that caused extra work for them, the issue is taken care of and I feel as a respected customer. It is very nice to be dealing with a company where customer service goes to 11.
So, anybody know what FedEx is doing to make the whole company to behave like that?
This is a small business. Mainly a one-person operation, although I do have some help every now and then and use subcontracting a lot. I do find the versatility of the tasks refreshing and fun. At some point, I’m implementing state-of-the art algorithms (some pieces of the LitePlacer software are from fresh doctorate thesis), and the whole software is a high-tech assignment. There are also mechanical design, although not recently. Lots of customer communication, which I enjoy. Also stock management, cash management, web site building, web shop management and sometimes, almost brain-less packing, floor cleaning and taking out the trash.
I like wearing many hats during a day. I also understand why some people with a different mindset would hate it. As a sample of my typical day, here is what I’m doing today:
I started the morning by taking and editing photos about some new additions to the web shop, wearing the hat of a photographer and graphics designer. Then, I added the products to the shop, playing as a web master for a while:
Then I talked to my accountant, being the financial officer for a while. I also talked to the print shop, being the marketing manager. I guess writing a blog post is kind of marketing, too. Then, I’ll pack and send a kit, being the warehouse worker. The office needs vacuuming at some point. In the evening, if I still have time and energy, I’ll concentrate improving the software as the head programmer, wearing the propeller hat. If the day runs out, I know where to start tomorrow.
I announced the availability of the kit for pre-order a few days ago, and of course, I wanted to get the word out to people that might be interested about it. I set up some Facebook ads and tried to target those to people that might be interested about it. I set the budget and let Facebook optimize the results. All looked good in the start.
A day or two later, to my horror Facebook told me that it has shown the ad in some areas of the world more than 20 times on average to single person. Auch! I certainly didn’t want to do annoy anybody, I wanted to impress, if possible. I tried to turn that down, but half a day later, I found out that it did not work. I was still spamming! I turned ads off and got professional help.
I still want to get the word out. However, nobody should now see the ad no more than once a day, if even that; just like any well behaving, self-respecting company should behave. The combination of an inexperienced advertiser and the cheapness of digital advertising can be a very annoying combination.
I spammed. I am deeply sorry and ashamed.
The CAD program I used first was the affordable Alibre PE. I took advantage of an update order and now use the not quite so affordable GeoMagic Design. The cheap but very useful PE equivalent is now Cubify Design. The mechanical design files should still open in Cubify; if not, please let me know! Feel free to adapt the design to your needs.
The screenshot and drawing utility I used in the documentation is called Greenshot.
Notable suppliers for parts are MakerSlide Europe, Inventables, MISUMI, RobotDigg and StepperOnline. Many smaller parts are sourced elsewhere, too. The steel plates are laser cut at Pelaser. The profiles were drilled and cut by Asesepänliike Teräsjänne Oy.
The software is written in C# using Microsoft’s Visual Studio Express 2013 for Windows Desktop. The video processing library used is AForge.NET.
Disclaimer: I have no other association to any of the companies mentioned other than a satisfied customer and I don’t get any compensation for the free advertisement – except a small comission if you use the Stylizer links when buying. Thank you for doing that, if you check it out and happen to like it! 🙂
They don’t have any special meaning. I’m using GeoMagic Design to design the machine and drawing the illustrations. I like that software very much, but there are some shortcomings. One of them is that the part trail color is not user-selectable. Some cases the default white works fine, some cases it doesn’t. When it doesn’t, I draw green trails manually.
Btw, the illustrations are screenshots, where I have manually added the balloons. The screenshot utility I’m using is called Greenshot; highly recommended! The green part trails and number balloons are drawn in Greenshot as well.
(I have no association with either of the companies, just a satisfied customer.)
A friend asked why am I not doing a Kickstart for this project? On one hand, it would make sense: A successful campaign provides financing to finish the development and parts procurement for the kit, and most importantly, gives some idea about the how many of you actually want to buy the kit. At this point, I have no idea if the kits will sell one a month or ten a day.
On the other hand, in this case Kickstart would not bring any customer benefits: I am not actually starving, and don’t need money for development. Time spent preparing for a Kickstart campaign is time not spent on development, so it would actually slow down things. From a customer point of view, there is not much difference between Kickstart and an order with a long delivery period – except higher price, Kickstart is not free!
So, I don’t bother with Kickstart. There just will be a longish delivery time after the beta series is sold. That time (I’m thinking two months or so) will give us an idea about the popularity of the LitePlacer and we can purchase parts accordingly, as well as allow us to streamline our processes to make everything happen smoothly. Most importantly, the time lets us sort out any issues that might surface when there are more than the few development machines in existence.