From Prototype to Production: 3 Things You Must Do There's no substitute for research. Your success depends on knowing what it costs to create your product and finding the best way to build it.

Creating new hardware products is hard. No surprise there. Every step along the way has its own pitfalls: idea, proof of concept, functional prototype, public launch, production, ongoing support. The process requires extended periods of research and development -- which in turn call for significant up-front investment.

Related: Watch This Guy Try, Fail and Ultimately Succeed to Build His Own Iron Man Suit

Entrepreneurs embarking on new-product development need the chance to experiment with something new and leave room for error.

But the downside is that all these steps make it easy for hardware startups to fail. Take the company behind the Lily Camera Drone. It raised $15 million in funding and received $34 million in preorders. But leaders over-promised on features as well as ship-date estimates. And, ultimately, the group closed its operations.

Another version of a drone, the Zano mini-drone, was Europe's biggest Kickstarter campaign. But it too shut down, after receiving $3.4 million from more than 12,000 backers. Poor planning and lack of experience led Zano's team members to overestimate their ability to execute their vision.

Companies that avoid these common mistakes excel in the hardware space. Form Labs, the $3 million Kickstarter recipient that developed desktop 3-D printers, successfully shipped its first-generation printer on time because the company launched its funding campaign well into the prototyping phase.

DJI, meanwhile, has dominated the drone industry because of smart strategies like locating its engineering offices and assembly facility in close proximity. This proximity created a tight feedback loop throughout the design and production processes.

Here at Double Robotics, we've made some mistakes and we've done some things right. We successfully shipped our first hardware product just nine months after its public debut. Our hardware product has been on the market for four years now, and we couldn't be more excited about our growth. It hasn't been easy; and there have been bumps along the way. Here are three crucial tips we learned:

1. Create a bill of materials with pristine accuracy.

A bill of materials (BOM) is a spreadsheet that details the complete cost of all individual parts and assemblies needed to create a single production unit. Your BOM also should factor in shipping for raw materials, import taxes and labor. Costs will vary with volume, but be sure to base your initial figures on the lowest expected production volumes. This is a safeguard in case sales come in lower than anticipated.

Estimate the BOM once the prototype is complete. At this stage, you won't have final production designs or quotes from your contract manufacturers. You'll need experienced mechanical and electrical engineers to envision the final product and estimate each part separately.

An accurate BOM is the foundation for your cost of goods sold (COGS) -- your total cost to get a single unit to the customer. Your COGS is equal to the BOM plus labor and shipping costs. Knowing your BOM and COGS allows you to set a sustainable selling price for your product: The difference between the COGS and the selling price is your gross margin.

Your margin will vary depending on your distribution channel. If you sell through retail stores, for example, you'll need a much higher margin than if you sell direct to customers. A margin in the 50 percent to 60 percent range is a good rule of thumb for startups selling to direct to customers.

Understanding exactly how much it will cost to build your product is the only way to ensure you won't lose money or be forced to raise prices after your product launches.

Related: Pricing a Product

2. Plan for the worst.

In any startup, resources are scarce, and time is of the essence. When estimating time lines, add 50 percent on top of your original estimate. With hardware, this is even more important, because you're moving around physical atoms instead of just bits online.

Here are a few areas that require extra planning:

• Production. Assume it will be very expensive and take much longer than you think to set up a contract manufacturer. It's always challenging to create systems for documentation, quality control and a repeatable assembly process. Parts often arrive from manufacturers with varying dimensions or specifications. It's extremely important to communicate your tolerance -- the range of specifications you can accept -- to suppliers.
• Shipping. Investigate all possible import taxes and fees, especially on inbound shipments coming from overseas suppliers. Will these arrive by air or by sea? Air is more expensive, but sea takes a very long time (and time is money). Some products cannot be shipped by air, and some products bring fees that outweigh the cost benefits of using an overseas manufacturer.
• Regulations. Build-in extra time for relevant government testing and certifications. These might include FCC Certification, UL for electrical and safety testing and CE testing for many countries around the world. Expect to iterate your design a few times before you pass these certifications.

Related: Why Small-Business Owners and Buyers Support Trump's Plan to Renegotiate Trade Deals

3. Create an extensible product.

When we first dreamed up the idea behind Double, we had a long list of features we wanted to include. We also knew we wanted to begin shipping as soon as possible. Creating an extensible product -- something that can be designed to accommodate new features or changes -- has allowed us to build on our product over time. Upgrades can come in the form of hardware, software or both. Think carefully about how these improvements will play out through the lifecycle of your product.

We've released three accessories, and all are backward-compatible to the very first Double units (built three years ago). Even our earliest customers are able to enjoy these new features. That helps us build strong relationships and a reputation as a company that supports its products.

Bonus tip: Consider keeping your final assembly in-house.

At our company, we're fortunate to have a low-volume, high-margin product, with final assembly in our California facility. Our engineers remain very close to the entire assembly process, so they can solve issues quickly and iterate with more freedom. Because everything revolves around time -- time to market, time to iterate -- it was crucial to ship our first product quickly. Keeping the assembly process in-house enabled us to ship six months sooner than we could have if we'd used a traditional contract manufacturer.

Related: 3 Road-Mapping Sins That Could Kill Your Next Product

If you forgo a contract manufacturer and retain all aspects of production in-house, you can iterate in the moment and conduct "just-in-time" manufacturing. You'll also have tight quality control and a narrow turnaround you might not otherwise be able to achieve. Keeping all engineers, tooling, processes and knowledge within your company's walls means you're able to truly invest in yourselves.

That's an important but often overlooked distinction. While the challenges of taking your product from prototype to production can be daunting, the final result is immensely rewarding and something to be celebrated as a team.