The smart camera is here…

Nest Labs, the company created by Tony Fadell and Matt Rogers in 2010 and bought by Google in 2014 for $3.2 billion, has just launched its fourth product, Nest Cam Outdoors, following on from its thermostat, smoke detector, and indoor camera.

The idea is a simple one: take the indoor camera and put it in a weatherproof casing with an attractive design, bidirectional audio that can be increased through strategically located speakers so that the wall works as a sound amplifier. What is really interesting about this though is not the product per se, but the strategic and accidental route that has led to its creation.

Nest bought Dropcam in June 2014 for $555 million. The company made WiFi-connected cameras that made it very easy to monitor interiors: you just placed the device in whichever room, connected it to the WiFi, and stored recordings of up to one month on the company’s own cloud space. Equipped with motion sensors, a microphone and speakers, it sent an alert to your smartphone when it detected noise or movement. The business model was equally simple: the camera cost $199 (there was a cheaper version with a narrower field of vision and lower resolution that cost $149), while a year’s storage on the cloud cost $199. Some 39% of clients that bought the camera decided to pay for storage as well.

But the acquisition did not go well, due mainly to differing leadership styles that led to mutual accusations between Tony Fadell, who said the team he had bought was not up to scratch, and the logical defense of the company’s founder Greg Duffy, all of which created an uncomfortable environment. After a while, Dropcam’s founder left, describing the whole affair as a big mistake. Meanwhile, the product had been rebranded as Nest Cam, while users of the Dropcam app were invited to download the Nest app.

So what is the really important thing that has been added to the Nest Cam Outdoors? It may not be visible in the design, but it is certainly what gives it the greatest added value: Google’s image recognition technology, based on machine learning. False alarms were one of the major weaknesses of the camera, so Google has combined its computerized vision technology and machine learning to estimate, with a reasonable degree of certainty, when movement is an insect walking across the screen, an animal or car nearby, or a person.

It’s also interesting the way the whole thing has come together: Dropcam provides technology through a difficult acquisition, Nest comes up with its usual excellent design and manufacture, and Google, which owns Nest (although the company has trumpeted its independence), contributes a layer of artificial intelligence that could end up providing an important differential value for a product in a highly competitive category.

There are any number of security cameras out there. In a market dominated by competitors that also offer a range of other security-related services such as alarms or guards, we now have an amateur offering cheap cameras that can be installed by the user. In between the two markets, there may be niches available, assuming the device doesn’t give false alarms and provides a reasonable level of security. In any event, what all this shows is that once again, the use of artificial intelligence makes a big difference.

(En español, aquí)