The history of a startup - we analyze other people's mistakes. Part 2
Good day, ladies and gentlemen!
In the first part ( here ), I began a story about how one startup was pierced with planning the availability of electronic components.
Now I’ll tell you how a startup got stuck with scheduling readiness for injection molds, and wasted time because of misuse of CAE (Computer Aided Engineering).
')
Read on
Initially, startups planned to get molded plastic parts in 3 weeks. But not so fast. In addition to directly producing forms, there is also a preparatory stage of work.
Standard mold manufacturing time is 5-6 weeks. But this is directly the time of production. In addition, the following steps go to the very beginning of production: prepayment from the client to us — interior design of the papers — prepayment from us plant-DFM (Design for Manufacturing) from plant-review and transfer of DFM- sending DFM to client-review DFM on the client side- DFM approval - the beginning of the production of forms. This chain can easily take 3 weeks or even more, especially if there are a lot of details and if there are any problems at each stage, or some stages are repeated (usually DFM).
Time, no doubt, can be reduced. Both from our side and from our subcontractors. But then pressure on both us and the foundry workers begins. We will speak directly, often with planning problems (both customers and our own, which really must be confessed) you will not be able to hang on the foundry workers. Plus it’s one thing to push if you have an annual order volume of only 10,000 parts, but the customer brings you a significant turnover. And another thing, if you will push because of the order in 1000 products per year from a customer who can be bent right after the first batch. And if you need a million parts a year, then the first client will wait, and the second one, and even more so, is at the bottom of the list of priorities :)
In general, the time required to prepare forms has increased dramatically from 3 to 9 weeks. That is the puncture they made when planning the readiness of cast parts. As I said, the guys decided to wait 2 months.
And now about how I pierced.
After some they showed us a promotional video. The product is funny, but I would not buy this for myself, because First of all, I don’t need this, and secondly, they want a lot for it.
After reviewing the video, I was asked how much the molds for the production of plastic parts could cost.
“I don't know,” I say, “I don't know how large the parts are, I don't know how many of them, I don't know how to group them in forms.”
- Well, you saw the video just now, tell me how much!
- I do not know, I need to watch 3D models and only then I can make preliminary estimates.
- Well, you have the experience, you saw the video, there are not many details there, and we don’t really need it. Approximately, to know how much to count.
Then I gave up and blurted out with an incomprehensible goal:
- Well, thousands of 75 yuan.
As it turned out a little later, the cheapest option was exactly 2 times more expensive, 150 thousand yuan. Having learned the real price for me, they tried to run into me, they say, you told us 2 times less! This is not what we expected!
All the same, in spite of the mistakes with the planning of electronics and mechanics, everything turned out not to be the way the startupers had planned. In fact, the start of sales they have occurred with a delay of 10 months from the planned date.
After our meeting, we corresponded for some time.
One of the questions was:
- And how can you reduce the weight of the product? I am now engaged in the modeling of injection molding and look to what can be reduced to a minimum wall thickness of parts. Material is PA (approx. By the author is polyamide, it is also sometimes incorrectly called nylon).
Vague doubts began to torment me, and I asked to send me a kinematic scheme.
“Why?” The question followed.
I think this point should be explained, because for many it is not obvious. An analogy from the world of software development — the guy started writing assembler inserts to optimize the bubble sorting algorithm, although we should start with a general design and algorithm selection (I hope I brought the correct analogy :)).
- Well, why, I asked. We will reduce the mass. The algorithm is simple, first we define the trajectories of motion, then we make up the kinematic scheme, then we look at which links we can throw out. The easiest detail is the one that does not exist :) Then we consider the acting forces. Then we’re looking at where concretely the forces are applied and begin to apply FEA (Finite Elements Analysis) to see what voltages these forces cause. Almost at the very end we select the material. Something like this.
I did not get the scheme, but I got a story about how he continues to optimize the details by modeling the casting. I explained to him that with the interpretation of the results one must be careful and at the moment that what he simulates there will in principle be wrong.
Again, I was asked a question, they say, why? I had to explain that it is of great importance how the details are arranged in the form, how many of them and how the gating system is designed. And since he does not know this and cannot design the form himself, the modeling of each individual part is meaningless, because then it may turn out that the parts are not spilled in the form. (I had a case where one part was spilled during the simulation, but four parts were not. In actual form, the parts did not spill either). But he did not listen to me and continued the fascinating meaningless exercise. Apparently, it was much more interesting to spend time on the meaningless pressing of the buttons and looking at the beautiful pictures in the program, than to solve the problem that he wanted to solve. Well, everyone decides what to do, depending on the goal, of course.
Another point was about the boards. They thought of using flexible boards to connect the control board with the drives. Well, I asked them, saying, why not a simple and cheap cable? The answer struck me, I did not even argue:
-We want our boards to be made as fast and cheap as possible anywhere in the world! And cables require manual labor! This is how it is written in the book, the advice of which I followed.
They would start as fast as possible, but no. The primary task is to think about how to produce boards anywhere in the world, because their start-up will instantly take off and production will need to be urgently organized around the world, otherwise they will not be able to cope with the demand. I didn’t tell them that the board requires FCT (Functional Test), ICT (In Circuit Test) may be required, besides, devices may be needed to fix the boards during transportation from operation to operation, since they are flexible. Tell about the car, where on one side you shove the connectors, on the other, the coil of the wire, and on the third you get the finished cable, I also did not. What is the point of proving something to people who basically do not hear you? This is probably because I consulted them for free, perhaps, if they were billed, then the attitude was different.
Some time later, when the guys disappeared for a long time, I climbed into their blog to read how they were doing there. Things turned out to be good - according to them, they change the material of forms from ordinary steel to high-quality steel and this takes a long time.
Again, they would start, because the delay is already more than half a year, but no, it is necessary to courageously urgently solve the problems that they themselves create from scratch. In my opinion, here again lack of experience in production and to the heap in project management affects.
What are the conclusions from the writing?
1. Knowledge is power! (From the first part, nothing has changed here).
2. Again, find out the details of what you do. Read books, read forums, see what problems people have, how they solve them, what questions they ask.
3. If the deadlines have already sunk strongly, and plastic parts are needed the day before yesterday, then the choice is obvious, but this is only if you clearly know what to do. The straight path is not the usual form, but aluminum. They are made of aluminum of the 7th series. A certain amount of money (2 times less than the usual form) and 2 weeks of time instead of 9. The durability of the form depends on the material, for polycarbonate or acrylic there are about 5000 products. The very thing for the first time to start.
4. And deducing personally for oneself, at first look at the details and only then give assessments.
In the first part ( here ), I began a story about how one startup was pierced with planning the availability of electronic components.
Now I’ll tell you how a startup got stuck with scheduling readiness for injection molds, and wasted time because of misuse of CAE (Computer Aided Engineering).
')
Read on
Initially, startups planned to get molded plastic parts in 3 weeks. But not so fast. In addition to directly producing forms, there is also a preparatory stage of work.
Standard mold manufacturing time is 5-6 weeks. But this is directly the time of production. In addition, the following steps go to the very beginning of production: prepayment from the client to us — interior design of the papers — prepayment from us plant-DFM (Design for Manufacturing) from plant-review and transfer of DFM- sending DFM to client-review DFM on the client side- DFM approval - the beginning of the production of forms. This chain can easily take 3 weeks or even more, especially if there are a lot of details and if there are any problems at each stage, or some stages are repeated (usually DFM).
Time, no doubt, can be reduced. Both from our side and from our subcontractors. But then pressure on both us and the foundry workers begins. We will speak directly, often with planning problems (both customers and our own, which really must be confessed) you will not be able to hang on the foundry workers. Plus it’s one thing to push if you have an annual order volume of only 10,000 parts, but the customer brings you a significant turnover. And another thing, if you will push because of the order in 1000 products per year from a customer who can be bent right after the first batch. And if you need a million parts a year, then the first client will wait, and the second one, and even more so, is at the bottom of the list of priorities :)
In general, the time required to prepare forms has increased dramatically from 3 to 9 weeks. That is the puncture they made when planning the readiness of cast parts. As I said, the guys decided to wait 2 months.
And now about how I pierced.
After some they showed us a promotional video. The product is funny, but I would not buy this for myself, because First of all, I don’t need this, and secondly, they want a lot for it.
After reviewing the video, I was asked how much the molds for the production of plastic parts could cost.
“I don't know,” I say, “I don't know how large the parts are, I don't know how many of them, I don't know how to group them in forms.”
- Well, you saw the video just now, tell me how much!
- I do not know, I need to watch 3D models and only then I can make preliminary estimates.
- Well, you have the experience, you saw the video, there are not many details there, and we don’t really need it. Approximately, to know how much to count.
Then I gave up and blurted out with an incomprehensible goal:
- Well, thousands of 75 yuan.
As it turned out a little later, the cheapest option was exactly 2 times more expensive, 150 thousand yuan. Having learned the real price for me, they tried to run into me, they say, you told us 2 times less! This is not what we expected!
All the same, in spite of the mistakes with the planning of electronics and mechanics, everything turned out not to be the way the startupers had planned. In fact, the start of sales they have occurred with a delay of 10 months from the planned date.
After our meeting, we corresponded for some time.
One of the questions was:
- And how can you reduce the weight of the product? I am now engaged in the modeling of injection molding and look to what can be reduced to a minimum wall thickness of parts. Material is PA (approx. By the author is polyamide, it is also sometimes incorrectly called nylon).
Vague doubts began to torment me, and I asked to send me a kinematic scheme.
“Why?” The question followed.
I think this point should be explained, because for many it is not obvious. An analogy from the world of software development — the guy started writing assembler inserts to optimize the bubble sorting algorithm, although we should start with a general design and algorithm selection (I hope I brought the correct analogy :)).
- Well, why, I asked. We will reduce the mass. The algorithm is simple, first we define the trajectories of motion, then we make up the kinematic scheme, then we look at which links we can throw out. The easiest detail is the one that does not exist :) Then we consider the acting forces. Then we’re looking at where concretely the forces are applied and begin to apply FEA (Finite Elements Analysis) to see what voltages these forces cause. Almost at the very end we select the material. Something like this.
I did not get the scheme, but I got a story about how he continues to optimize the details by modeling the casting. I explained to him that with the interpretation of the results one must be careful and at the moment that what he simulates there will in principle be wrong.
Again, I was asked a question, they say, why? I had to explain that it is of great importance how the details are arranged in the form, how many of them and how the gating system is designed. And since he does not know this and cannot design the form himself, the modeling of each individual part is meaningless, because then it may turn out that the parts are not spilled in the form. (I had a case where one part was spilled during the simulation, but four parts were not. In actual form, the parts did not spill either). But he did not listen to me and continued the fascinating meaningless exercise. Apparently, it was much more interesting to spend time on the meaningless pressing of the buttons and looking at the beautiful pictures in the program, than to solve the problem that he wanted to solve. Well, everyone decides what to do, depending on the goal, of course.
Another point was about the boards. They thought of using flexible boards to connect the control board with the drives. Well, I asked them, saying, why not a simple and cheap cable? The answer struck me, I did not even argue:
-We want our boards to be made as fast and cheap as possible anywhere in the world! And cables require manual labor! This is how it is written in the book, the advice of which I followed.
They would start as fast as possible, but no. The primary task is to think about how to produce boards anywhere in the world, because their start-up will instantly take off and production will need to be urgently organized around the world, otherwise they will not be able to cope with the demand. I didn’t tell them that the board requires FCT (Functional Test), ICT (In Circuit Test) may be required, besides, devices may be needed to fix the boards during transportation from operation to operation, since they are flexible. Tell about the car, where on one side you shove the connectors, on the other, the coil of the wire, and on the third you get the finished cable, I also did not. What is the point of proving something to people who basically do not hear you? This is probably because I consulted them for free, perhaps, if they were billed, then the attitude was different.
Some time later, when the guys disappeared for a long time, I climbed into their blog to read how they were doing there. Things turned out to be good - according to them, they change the material of forms from ordinary steel to high-quality steel and this takes a long time.
Again, they would start, because the delay is already more than half a year, but no, it is necessary to courageously urgently solve the problems that they themselves create from scratch. In my opinion, here again lack of experience in production and to the heap in project management affects.
What are the conclusions from the writing?
1. Knowledge is power! (From the first part, nothing has changed here).
2. Again, find out the details of what you do. Read books, read forums, see what problems people have, how they solve them, what questions they ask.
3. If the deadlines have already sunk strongly, and plastic parts are needed the day before yesterday, then the choice is obvious, but this is only if you clearly know what to do. The straight path is not the usual form, but aluminum. They are made of aluminum of the 7th series. A certain amount of money (2 times less than the usual form) and 2 weeks of time instead of 9. The durability of the form depends on the material, for polycarbonate or acrylic there are about 5000 products. The very thing for the first time to start.
4. And deducing personally for oneself, at first look at the details and only then give assessments.
Source: https://habr.com/ru/post/292646/
All Articles