Prototyping failure

A few more years, a couple of adventures in restaurant and bakery businesses, and I was again engaged in the project for a machine.

After reconnecting with a friend, a brilliant inventor with whom I developed pharmaceutical machines in the nineties in Italy, I had been invited to join the team in his start-up backed by a solid Italian pharmaceutical and co-financed with a public grant to develop a new technology to inject blow-mould syringes made in one single piece, pre-filled and sealed within a single machine.

It was a very ambitious project but my friend already had the tricky details of the syringe on paper and with our combined experience and skill we were very confident.

When I joined the team a decision had already been made to locate the project in China as apparently there were quite some advantages in terms of costs and prototyping speed.

So, when I escorted my friend and his partner for a visit in Guangdong they got enthusiastic about Ronghui, a sub-district of Shunde, Foshan, where apparently mould machining shops were springing up like mushrooms after the rain.

They rented a workshop in a mould industrial park very close to the fast railway connecting it to Guangzhou down-town in less than 1 hour, opened the WOFE, Wolly Owned Foreign Enterprise, and got the basic equipment ready as well as a standard injection blow-moulding machine to use as a test bench for the mould before building our own machine.

Learning curve

The plan was very ambitious. We expected to have the injection mould ready for testing in 2 months for the first iteration and then start working in parallel on the blow moulding part and arrive to a functional prototype of the whole moulding in less than 12 months.

We were in the mould paradise, we thought, so all should have been just fine and fast.

Of course we met our troubles with the design of the mould and got a few patents registered for the solutions we invented, but soon enough we sent out our production instructions to the machining shops.

The magic was happening: we were getting our mould done by next Monday, machined and tempered!

Well, all threads in the holes were only 5 mm deep, the overall size was off of a couple of mm, all corner were oddly rounded, and the stainless steel was rusting.

I called the stainless steel supplier asking about the rusting German imported stainless steel and the candid answer was: “well, you did not ask for the one that does not rust. We also have that one. Which colour you prefer?”

My friend came back from the machining company, our neighbour, with a desperate face and a similar story of incompetence and approximation. The workshop even had no one gauge available to check measures.

We started suspecting that our Heaven of the moulds was going to be more of a purgatory.

Weeks were passing and one after another unusable moulds were piling up.

But every learning curve sooner or later tends to knowledge so we eventually found the ultimate machining workshop. They had everything, including a laser gauge. Our original materials arrived from Germany and we went to the workshop and sat there a whole day to assure every step was executed according to industry standard.

At ten in the evening, skipped lunch and dinner, the mould was finally complete. Only one fixing slot remained to drill when, against my advice, we left.

Three days later the mould was sitting on our work bench ready for assembly. By 8 that night we finally had it on the injection machine and we started fixing it with the bolts in the fixing slots. 3 done, one to go. But it never happened. The hole was off by more than 2 mm.

Machining with an artist’s touch

We checked the 3D drawings, we called the designer in Italy, finally we took a gauge and measured the fixing slot. It was obviously wrong.

What happened is very simple and obvious.

• The worker who was in charge the night we left our mould in the CNC machine was the smartest in the shop.
• He was tired and he wanted to go home.
• He noticed that only one hole was left to complete.
• He knew that to drill holes you need a drill and a drill is faster than a CNC.
• He stopped the CNC machine, took a large hand held drill and drilled the hole, but the drill was big and interfered with the mould, so he had to drill the hole 2 mm off.

The mould never formed a syringe, my friend’s machine was never built, the Pharmaceutical company had its returns anyway thanks to complex financial operations, I was paid for learning again a few lessons and my friend learned that my 20 years of experience in China were better used if he listened to me than having me assembling moulds.

The lesson I learned

• Make informed choices where to develop your project in terms of business logistics and culture
• Make sure that everyone is clear and consistent about what their returns are
• Better to say “no thanks” sometimes