Time to read: 5 min

Software and hardware engineers alike have heard some form of this advice: “Build lots of prototypes; iterate as fast as you can.”

For those of us in the hardware industry, these prototypes can cost a lot of time and money. So, while we should strive to iterate quickly, we must also get as much information as possible out of each round of prototypes to inform improvements in the next iteration.

This is the purpose of hardware product testing. To determine how a piece of computer hardware prototype behaves. 

If you’re an Industrial or UX Designer, you might be most familiar with user testing. If you’re a Mechanical Engineer, you might picture things being stressed to their breaking point (and aren’t failures oddly satisfying?). Electrical and Firmware Engineers might think of a bed-of-nails fixture.

All the above are part of a comprehensive test plan for hardware products—different tests are beneficial at different stages and each test is an indispensable part of the hardware development process.

This introductory article will give an overview of the different types of tests to consider at each major stage of hardware product development and share some best practices for building a testing schedule and testing methods to help you get started. While the main goal of this article is to review hardware testing methods, software testing may be briefly discussed.

Photo by Christian Wiediger on Unsplash

Early-Stage Testing

The first step to testing is to ensure you have an adequate hardware testing process/plan with appropriate hardware testing methodologies and testing hardware. Always consult with the end-product requirements which may be flown down from the customer, specifications, and/or industry or legal regulations. Keeping these things in mind will ensure your hardware product testing strategy covers all of the required hardware design and testing methods as well as hardware checks that must occur before hardware is released to production.

A lot of hardware tests can be performed with even the crudest prototypes. There are 3 major test types you’ll want to build into your schedule during early-stage development:

  1. Usability tests to help refine the industrial design of the product
  2. Material and finish tests to assess different options for durability
  3. Critical components testing to analyze different mechanisms in isolation before a complete looks-like, works-like prototype is ready.

Here’s an example of what some early-stage hardware test plan overviews might look like for a mobile phone product for kids:

Examples of early-stage tests

These tests will give you the feedback you need to iterate on your designs and ultimately build a working prototype.

Pre-Production Testing

As your prototypes become closer to the real thing, your testing methods need to evolve as well. Testing hardware with the pre-production build methodology (EVT, DVT, PVT) gives you prototypes that are much closer to the final product vision.

With these high-fidelity prototypes, you can finally do system-level stress tests, environmental tests, and life tests. Compliance certification tests (e.g. FCC, CE) can be done at this stage as well.

Here’s an example of some of the proto-build stage tests you might do:

  1. Thermal shock / thermal cycling
  2. High temperature, high humidity operation
  3. UV exposure
  4. Salt spray
  5. Household chemical reaction
  6. Water spray for IPX4 certification
  7. Water immersion for IPX7 certification
  8. Dust ingress
  9. Electrostatic Discharge (ESD), contact and air
  10. RF emissions
  11. 6-Face Drop
  12. Tumble
  13. Vibration
  14. Abrasion
  15. Switch cycling

To develop these tests, you’ll want to work in partnership with your contract manufacturer. The purpose of these pre-production tests is to perform all hardware qa testing, functional testing, product validation testing, and manufacturing validation. 

Mass Production Testing

Testing doesn’t end as mass production begins. On the contrary, testing needs to continue to happen for as long as the assembly line is still running. Hardware QA testing continues throughout production and is typically less involved than pre-production testing; rather it functions as a product verification of the hardware as it is produced. 

Testing during mass production has a very different purpose: to ensure that the 10th and 50,000th product coming off the assembly line are the same (sameness being a range, but that’s another topic for another article!).

Note that some tests can and should be performed on 100% of the products before they leave the factory, but destructive tests cannot, so a statistical approach needs to be taken. Always consult with the requirements for each hardware product before determining which production hardware tests must be performed on each lot or specific quantity of hardware components produced.

Test Schedule Example

As you plan out your product development schedule, it’s crucial to budget ample time and manpower for testing like any other development process.

At the very least, make a list of all the outsourced tests you wish to have done before EVT, then use this list to get quotes and lead times so you can plan ahead.

Here’s an example of what a portion of your hardware testing schedule might look like:

 sample hardware testing schedule

Note that it’s important to build testing lead times into your schedule in addition to time needed for design changes if the tests uncover something that needs to be improved.

Here’s are some of the major considerations to remember when building your test schedule:

Plan the Number of Finished Goods

For some tests, complete units are not required so having a test plan can help you budget the number of finished goods needed from each build. For example, UV tests and abrasion tests can be done with just the cosmetic components.

Regardless of whether the test requires complete units or individual components, it’s important to provide enough test samples so you have confidence in the results.

Consider the Requirements of Custom Fixtures

A complete suite of standard environmental tests can cost thousands of dollars. In addition, you might want to build custom fixtures that simulate the unique use cases of your product, which cost money and significant man-hours.

In my experience, fixtures make great intern projects as they are self-contained and short-term.

Don’t Rely on Your Contract Manufacturer for QA

Every software team needs dedicated QA (quality assurance) engineers and it’s no different for hardware teams.

It’s quite common for small engineering teams to not have hardware QA resources during the design phase, but as mass production gets closer QA is an indispensable role in the manufacturing operations team.

Don’t rely on your contract manufacturer to take complete responsibility of QA, as there is an inherent conflict of interest. A mass production test plan should be a work of collaboration between in-house QA and the contract manufacturer.

Taking the time to test designs early in the product development process always saves time and money later. Following these hardware testing basics will pay off when hardware production begins.

On the other hand, skipping tests during the design phase carries the risk of a huge reset when problems are uncovered during the pre-production stage and when the cost of change is high. In the worst-case scenario, paying customers could discover flaws that should have been caught well before the factory started churning out hundreds of units per day.

Don’t tempt Murphy’s law: design a test plan and dedicate ample resources to testing so your final product meets your, and your customers’, standards.