In the past few years, the automotive industry has been shifting gears to make cars that are Autonomous, Connected, Electric, and/or Shared (ACES). So much so, that according to a McKinsey & Co. report from March 2019, $210B has been invested in mobility companies focusing on ACES. And this number has only continued to rise.

For those of us who have made IoT our bread and butter, the ACES trend is no surprise. Every industry has been affected by the proliferation of connected devices, and the automotive sector is among the largest. In the context of automobile production, this becomes all the more important. Since Henry Ford’s great success in mass manufacturing automobiles 100 years ago, the automotive industry has carried both the burden and prestige of being considered one of the most efficient and forward moving industries. Here’s a fun piece of anecdotal evidence: in 2013, Toyota literally donated efficiency (in lieu of cash) to the Food Bank for New York City.

Despite the automotive industry’s efficiency, it’s struggling to scale the production of ACES cars in a timely manner. Ford figured out how to mass manufacture the mechanically complex internal combustion engine (ICE), but today’s automotive leaders have an even more difficult task. Today, the industry’s leaders have to figure out how to integrate high-tech IoT components into cars in a way that’s safe enough to be approved by the public, cool enough to market to picky consumers, and also make production scalable. Scalable production is the topic of interest for those of us who manufacture things.  

One big case study is Tesla. Tesla has been struggling to reach its ambitious production goals. There’s very little doubt that Tesla is capable of designing beautiful cars, but manufacturing these cars en masse is especially difficult. Why might this be the case? Popular news might lead you to believe that it’s related to battery issues. This seems reasonable since EVs rely on batteries in a way that ICE cars don’t. But as Musk points out, the delay is not unique to Tesla:

Tweet from Elon Musk, Emperor of EVs, space exploration, and sassy tweets.

Thus, we (those of us at BCD) think there’s another source of problems that has not been receiving enough coverage. By no means is there a single magic solution to ALL of the scaling issues related to the automotive industry. Instead, we think it’s time to emphasize a topic that’s already significant to insiders of the automotive industry. Testing. Yes, testing.

Embracing vehicles that are autonomous, connected, shared, and electric requires the integration of IoT components in a way that is unprecedented. While car makers like Tesla and Ford, are doing the bulk of the car’s design, they outsource the production of the circuit boards to CMs. This means that in a modern car, there are numerous components (some more important than others) that may have been designed and produced by multiple companies. These different companies probably have very distinct methods for product design and testing. Ultimately, all these parts have to be tested separately, as well as together to make sure everything works together to create a seamless experience for drivers.

Production line testing is necessary to ensure consistent quality in every product being manufactured, but it’s even more heavily scrutinized when you’re producing cars. So much so that there’s a whole set of events regularly staged around the world centered around automotive testing. There are already brilliant systems for checking the traditional parts of an automobile (i.e. engines, exhaust, wheels, and safety), but there are few established systems in place to test IoT devices. IoT devices are trickier to test because you’re not testing just the physical parts, but the firmware that makes the device work. Unsurprisingly, production testing for products with IoT components (like autonomous cars) can be quite cumbersome. It’s also often the bearer of bad news when you’re trying to scale production. When you have IoT components in the mix, a meticulous design and perfect prototype do not always translate to timely mass production. The amount of time required to setup production testing and run all the individual tests is already a major cost. But if something goes wrong, the problem must be diagnosed and solved ASAP! And in the worst cases production might have to be completely halted to address the issue.

Given today’s globalized economy, it’s common for development teams to be based far away from production teams. This means an engineer who designs a product and tests the prototypes may be on a different continent from where the product will be manufactured. And when something goes wrong, it’s usually difficult to remotely diagnose because there’s not an established system to let development teams know exactly what’s happening on the production line. Even video surveillance comes short because it cannot detect the invisible errors in equipment and devices.

Production testing is a critical part of bringing a product to market, but it also comes with a lot of hassles. That’s why Blue Clover Devices is proud to have designed the Production Line Tool (PLT), which is a cloud native test automation device that helps scale production testing of IoT products. The PLT is the first affordable automated test device. Automated testing means production testing is no longer a hurdle in scaling production. Meanwhile, a cloud native platform means engineers are able to remotely deploy firmware onto devices as well as view accurate test results, making it easier to diagnose a problem on the production line without being there in-person.

This may be bold to say, but at BCD we believe every creator strives to make excellent products. In order to consistently deliver excellence, we have to embrace production testing. To make it scalable, we have to embrace automated testing. Thus, to make excellent autonomous cars in a scalable way, we need autonomous testing. That’s why we made the PLT.

***Minor edits made on July 1, 2019 to address technical points***