Over the last 18 years, I’ve had the opportunity to experience the manufacturing journey first hand. When I joined iRobot to set up and lead the manufacturing team for Roomba, we needed to figure out all of the hardware from scratch. Just making a motor spin required designing an H-Bridge, picking out components, and writing custom firmware. This process took about three weeks.
It’s incredible to see how things have evolved, and how much opportunity for change is still on the horizon. Getting from an idea to a compelling prototype is easier than ever before and while manufacturing is becoming more accessible, it’s still tricky to navigate. I see the next wave of innovation in hardware being focused around making manufacturing more accessible.
Successfully manufacturing products at scale is inherently complex — it requires:
- Diverse set of skills, including project management; technical expertise in areas such as injection molding, die casting, sheet metal, PCB assembly, test fixtures, etc.; estimation and forecasting skills; and vendor selection and management.
- Ability to effectively communicate, often with team members and partners separated by different time zones, languages and cultures.
- Understanding of the long-term impact of decisions made early in the manufacturing process. For example, balancing your payment terms with your factory and with your customer can make or break your company.
- Tools to model cash flow and break-even quantities, and to calculate Cost of Goods Sold (COGS), and capital expenses (i.e. the cost of tooling and fixtures).
- Ability to select appropriate fabrication methods (i.e. 3D Printing, Machining, Injection Molding, etc) with cost and quality in mind.
- Strong and trusted network of factories, suppliers and partners that often takes many years to develop.
Early Decisions Cast Long-Shadows
Decisions made early in the manufacturing journey are critical, as they often cast long-shadows that can ultimately define the difference between success and failure. Unlike software development, it is much harder to iterate and optimize hardware products the closer you get to production.
When we were preparing to ship the first batch of 5,000 Roombas, we decided to test a few units for longer than our standard 5-hour tests. Much to our horror, after about 8 hours of run-time most robots developed a fatal flaw and started spinning in circles. When we took apart the units, we discovered that dust was clogging up the optical encoders. With the shipment deadline just 36 hours away, we couldn’t change the tooling and had very limited options to fix the problem. Though we were able to avoid disaster, after 5 years of hard work we almost failed at the finish line.
Cost, Quality & Schedule
With any complex task, it helps to break it into simpler pieces. We’ve found it effective to break down manufacturing into the following:
These “points of the manufacturing triangle” are all interrelated. For example, if you increase the cost of a product, you can often improve the quality and accelerate the schedule – using a more expensive motor may require less software development and provide a shorter lead time. However, the art comes in finding the right balance between all three points. In this example, if the cost is increased too much, backers may not be willing to in contribute to your campaign.
Phases of Hardware Journey
It’s helpful to think about the hardware journey as a series of phases. Doing the appropriate due diligence in the early phases will help you to set yourself up for success during the later, and more costly, parts of the journey.
This is where the light bulb goes off and the great idea happens. Lots of whiteboard scribbling, popsicle sticks and hot glue. Start putting together documents describing the problem to be solved, and some potential solutions. Talk with people to get their take, and build up confidence that it’s worth investing more time to build up a prototype.
These can be “looks-like” that are aesthetic models with limited functionally; “works-like” that demonstrate the basic features, but often in a non-realistic housing; or “works-like looks-like” that combines them both. Most of the time, prototypes are made by 3D printing CAD files and using Arduino or Raspberry Pi. Through multiple iterations, the prototype is refined to a point where it looks and works as expected, but is a custom assembly without design for manufacture.
Once you’ve built a few solid prototypes, and are confident that your design meets the requirements and goals that you specified, you should start work on building a community and narrative for your funding campaign. You still have a lot of pre-production work ahead of you to prepare for manufacturing before you launch.
At this point, start figuring out how to scale the prototype into higher volume production. If you are using injection molding, start adding “drafts” and “rounds” to the part files, and incorporating assembly features such as self-locating screw bosses, snap clips, ultrasonic welding features, etc. From these files, it’s possible to calculate the COGS, capital requirements and schedule.
Once you’ve developed a solid estimation of how much capital you will need to manufacture your product you can figure out the funding goal for your Kickstarter campaign. You’ll need to have a clear sense of how much it will cost to produce each unit to determine the reward costs. Be sure to include shipping and some contingency, as issues will always pop up. Also, consider what is the maximum quantity that you can produce, not just the minimum.
Before you can start manufacturing, you need to select a factory. This is the Request for Quote (RFQ) process which should follow a structured approach to evaluate multiple contract manufacturers (CMs) in such a way that they can be compared objectively and a decision made on the best one to partner with. Once this decision is made, the teams work closely together to refine the design for manufacturing and assembly to match the tools and capability of the factory. Tools are then cut, and the product progresses through manufacturing milestones including. Engineering Validation Testing (EVT), Design Validation Testing (DVT) and Production Validation Testing (PVT).
Once the product has been produced, it must still be delivered to the customer. This involves a Logistics team to fill out the appropriate shipping paperwork and transport the goods from the factory to the customer.
Think of this as a high level map to help you go from an idea to putting your product in backers’ hands. This is just the beginning. In future posts, we’ll drill down deeper into each phase and provide tips and tools to help you succeed. Our goal is to help creators understand the steps required to successfully bring their product to life ahead of launching their campaign. We are excited to provide creators with the ability to better prepare for delivering to their backers after the campaign ends by preparing before it launches.