Fail fast, fail often is the golden rule of the mythology of startups. It suits software developers perfectly. When code breaks, patch and push update and re-try. The price of loss is practically nil.
However, to the hardware engineers, who happen to be the ones constructing drones, robotics, future vehicle designs, or performance sports gear, this piece of advice sounds like a sick joke. Failure is costly in the physical world. And it is even more costly to do it with highly developed composites.
This gives rise to an effect called the Mold Trap. It is a certain stage of the development cycle where innovation is dead in the water, strangled by the restrictive economy of old world manufacturing.
The High Performance Paradox
The following is the situation: A drone engineer creates a new revolutionary drone chassis. The chassis needs to be extremely light and extremely stiff in order to realize the required time in the air and payload they can carry. Metal is too dense; the common plastic is too fragile. A composite is the only material that can be used.
The design is tested by the engineer using CAD software. It looks perfect. However, to determine whether it will really pass a crash landing or high-G manoeuvre, they must have a physical component tested.
Here they fall into the trap. With the conventional method, they require a mold to prepare one unit of testing. That mold is usually made of aluminum or steel to resist the heat and pressure of the curing composite. That one mold could be sold at 15000 and can be machined within six weeks.
Now before the engineer is an overwhelming decision:
- Gamble: put up the $15,000 and hope that the design works on the first attempt (it never does).
- Compromise: Try a cheap plastic one that does not act in the same way as the end product, and is going to provide no useful data.
- Stall: Kill the iterative process, test the design when it is frozen.
The Mold Trap compels genius engineering teams to be conservative. They do not take geometry and aerodynamics to their extremes but come up with safe and boxy forms, which are easier to tool. Innovation fails to die due to bad ideas, but it is the cost of validation that is prohibitive.
Escaping the Valley of Death
The Mold Trap solution is to separate the performance of the part and the permanence of the tool.
The silent revolution is a revolution in the development of high-performance hardware. The sector is no longer following a waterfall cycle of Designing, Tooling, and Testing, but rather a more flexible work flow. This has been driven by technologies that enable the production of composite parts with the lead time of six weeks like a steel tool.
The industry is being transformed by two main ways of escape:
1. Additive Tooling (Printing the Mold)
Engineers are now working with high-temperature industrial 3D printers to machine the molds themselves rather than machining a block of aluminum. Such molds are printed using heat-resistant polymers that are resistant to the autoclave or oven curing process.
It is no longer necessary to spend a month and pay $10,000 to get a mold, which could now be printed overnight at a cost of $400. In case the final component exposes a weakness of design, the engineer just makes minor adjustments on the CAD file and prints another mold the following day. This enables five or six design cycles in the time that it took to design one.
2. Direct Digital Fabrication
Worse still, the eradication of the mold is more radical. With the new manufacturing technologies, it is possible to deposit continuous fibers directly. It is now possible to lay strands of carbon reinforcement where the stress loads are needed with machines that follow the complicated curves of the design, and do not need to be shaped on the back side.
The Return of “Fail Fast”
The elimination of the financial penalty of tooling will enable the hardware teams to finally embrace the fail fast mode of their software counterparts.
Take the case of a suspension arm of an electric racing car. In the ancient design, the team would over-engineer the component to make it safe leading to a heavier part. Under the new model, they have the opportunity to plan three variants, one of them being a conservative, another aggressive, and the other one radical. They are able to print all the three within several days and painfully experiment with them on a rig.
They are in a position to shatter things. And because they are able to break things, they get to know where the real boundaries of the material are.
It is something that transforms the psychology of the design room. Increase in the cost of a mistake to the level of negligible combined with decrease in its cost to career-ending encourages creativity. Designers begin exploring the use of organic geometries which are generative and would be a nightmare to machine yet simple to print. They begin to optimize in the mass and aerodynamic chapter instead of optimizing in the constraints of a CNC mill.
The Future is Agile
The times of measuring ten times, cut once are evaporating. The currency of the day in a hyper-competitive international market is speed. The firms that emerge victorious in the decade to come will not always be the ones that have the largest budgets, they will be the ones that can fastest iterate.
Through adopting carbon fiber prototyping, organizations would be able to seize the digital to physical gap. They can break out of the Mold Trap and make the risky venture of hardware development a fast, repeatable science. The outcome is innovative, light and strong products that are introduced in the market months before the others.