Share This article

Google’s Project Ara first debuted more than a year ago, promising a landmark achievement in mobile computing — an upgradeable, modular smart phone that end users would be able to customize with their own hardware loadouts for various projects or uses. It’s been nearly two years since Google first unveiled the earliest sketches for what became Project Ara, and if the company had kept to its previous timetable, it should be gearing up for a field test in Puerto Rico by the end of the year. Unfortunately, that’s not going to happen in 2015. The Puerto Rico test has apparently been nuked, and Google will take Ara somewhere else for its pilot program. It’s also delaying the launch of the program until 2016.

The economic upheaval currently plaguing Puerto Rico could be responsible for Google’s decision to seek another test site. The island, which is classified as a US territory, recently defaulted on a bond payment worth some $58 million after decades of poor job growth and limited opportunity left the nation without enough capital to continue paying its bills. The relationship between the US and Puerto Rico is complex, to put it mildly. The island is officially claimed by the United States, but receives only a fraction of the social services and tax advantages as full-fledged states. Google may have seen which way the wind was blowing and chosen a market with better prospects for testing its new modular technology.

The upcoming third-generation device

The fact that Project Ara has been pushed back to an undisclosed date in 2016, however, suggests that Google is still trying to fix some fundamental issues with the device. It’s not hard to guess what those might be, since the company has previously admitted that power consumption has been problematic, and the modular components wear out relatively quickly as the data connectors are continually being swapped in and out of the smartphone. Spiral 3, which is supposed to add a new type of modular connector, improve battery performance, and boost RF performance, was supposed to be out this year. Up to 30% of Project Ara’s power consumption is spent moving data between the various modules, according to Re/Code. That’s not actually that surprising, since one of the advantages of tightly integrated SoC packages is precisely that you don’t have to move information very far across the chip. With Ara’s modular approach, data may have to travel across the entire phone body, whereas a conventional smartphone design would integrate sensors as tightly as possible to minimize such power consumption.

I’ve always liked the idea of Project Ara, personally, but without information on exactly which modules will be supported it’s difficult to tell how useful it would be in real life. I’m not willing to take a 30% hit to already terrible smartphone battery life, and while I love the idea of being able to add specialized peripherals to my smartphone, the truth is, I rarely have a problem with the device being outdated, even though I only upgrade every other year. The rapid-fire improvements to the smartphone industry that we saw from 2006 – 2012 have slowed in recent years. It’s not that Samsung and Apple don’t keep iterating, but that performance hit “good enough” several cycles back.

We’ll have to wait for 2016 to see what kind of improvements Google has made to the devices — maybe by then we’ll know if the hardware is actually going to come to market, or simply be shelved as another failed experiment.