Google IoT: The grand vision

Throughout its 20 years of existence, Google’s strategy has evolved around its search engine. Slowly but steadily, the company is pushing for growth in areas beyond online advertising. Cloud computing and artificial intelligence are two new topics Google’s holding company Alphabet is now pursuing as part of its Google IoT strategy.

When in 1996, the market for search engines was al ready dominated by en Google started-up – the likes of Yahoo, Altavista, and Lycos. Google’s founders Sergey Brin and Larry Page studied these engines in detail and were able to come up with a much superior technology that ranked websites through backlinks rather than via search-term occurrences. This new approach became, and remains, the source of Google’s success. The online search and advertising business has become mature, leading Alphabet/Google to explore growth avenues outside search – the Internet of Things being one of them. “The internet will disappear,” proclaimed Google’s chairman and ex-CEO Eric Schmidt in 2015 at the World Economic Forum in Davos, Switzerland. He was referring to the Internet of Things which he believes will lead to “interactions with devices that will be totally seamless.”
To act on this vision, Google is enhancing its cloud services and artificial intelligence capabilities, developing a number of IoT standards while working on various applications both in the consumer and enterprise IoT sectors. Many of its projects are based on open source frameworks and are performed in collaboration with other market participants. Google hasn’t been particularly quick in its quest to become a leader for the Internet of Things. In fact, other giant software firms like Amazon and Microsoft have gotten to market much quicker with their IoT-specific cloud adaptations. They have also made a more aggressive marketing efforts. Microsoft’s online presence, for example, is filled with numerous IoT success stories, videos and demos. Google lacks this breadth of showcases.

It seems Google is taking the approach of the ‘smart follower’ slowly building its capabilities but waiting for the right moment to strike – much like the company did in its early days when other search engines were dominating the market.

Where Google stands today

Dutch conglomerate Philips is a pilot customer of Google in the Internet of Things. Philips’ innovative line of app-controlled Philips Hue lamps is one of the early IoT success stories, built on top of Google’s infrastructure and using the full suite of its Cloud Platform services. Google’s Container Engine, powered by Kubernetes, and its Compute Engine handle over 200 million transactions every day, including at least 18 million remote lighting commands. The app engine is built on Google Bigtable, thereby managing data with an extra layer of built-in encryption, which eased Philips’ concerns about protecting customer information, according to Google.

Google has shown that a firm can be the best without being the first.

Overall, the company’s IoT strategy looks highly compelling, especially when one considers that several lower-level IoT platforms, such as Particle.io, Telit, and Sierra Wireless, are hosting their data on the Google Cloud. At the same time, outside the Internet of Things, the company is already managing trillions of data transactions for customers like Disney and Coca-Cola. However, in addition to being slow to market, Google has one other striking weakness in the Internet of Things. It lacks industrial use-case experience that goes beyond home and consumer applications. According to a report by consultancy firm McKinsey, 70% of the value in the Internet of Things will be captured by non-consumer business-to-business solutions. Large industrial players, including General Electric, ThyssenKrupp, Siemens, Schneider Electric and Rockwell Automation, are working with Google’s main competitors Microsoft, IBM, and Amazon for their stored cloud data and to access corresponding big data analytics.

Let there be light: Google’s Container Engine handles over 200 million transactions a day, including at least 18 million remote lighting commands.

Google will have to make an extra push if the firm wants to dominate the industrial Internet of Things in a similar way it is dominating online search today. The Internet of Things is growing fast but Google’s previous late entry into the search engine market showed that a firm can be the best without being the first. Perhaps history will repeat itself.

The current Google IoT strategy can be broken down into six elements – with each element enjoying its own dynamic. All six taken together presents a picture of a fairly compelling offering for the Internet of Things.

6 elements of Google IoT strategy

Consumer ioT products

Google’s acquisition of the Nest smart thermostat in early 2014 was seen as the start of Google’s attempt to dominate the smart home of the future, and it certainly raised mass awareness of IoT. But, while Google has expanded the product line and introduced the ‘Works with Nest’ proprietary standard, sales have fallen short of expectations. Nest’s founder Tony Fadell stepped down in 2016. The new flagship consumer product is Google Home, a voice-activated smart hub and speaker, launched in November 2016. The product is the first real competitor to Amazon’s Echo which launched almost 18 months earlier, with sales going into the millions. Google was able to build its product on top of its existing Assistant, which provided the instant ability to have natural, contextual conversations – an important feature which could give Google Home an advantage over Amazon Echo.

Cloud platform

A core element of nearly any IoT solution is some form of private or public cloud where data is stored and powerful analytics can be performed. Amazon Web Services (AWS) and Microsoft Azure are leading platforms in this field. Both companies have heavily amended their existing offerings to offer new products that facilitate the connection of IoT devices – AWS IoT Platform and Microsoft Azure IoT Suite. Google is starting to do the same and has teamed up with Intel to offer a better connection to Google Cloud for Intel’s IoT edge devices for retail and manufacturing. Several other partner companies are developing IoT-specific cloud add-ons. One example is Agosto which has developed a custom message queue telemetry transport (MQTT) broker for Google’s Compute Engine which provides virtual machines to enable scaling from single instances to global, load-balanced cloud computing.

Artificial intelligence/big data tools

Google’s CEO Sundar Pichai says his firm is “making a big bet on AI and machine learning.” This is an area in which Google may actually have an edge compared to its competitors – after all, its core business has always been about making sense of massive amounts of data. The company is investing heavily in machine-learning research and analytics capabilities. Its open source TensorFlow machine learning libraries are already used in numerous IoT and non-IoT applications. Aircraft manufacturer Airbus, for example, is using Tensorflow in conjunction with cloud computing tools such as Google Cloud Dataflow and BigQuery to automate the process of detecting and correcting satellite images that contain imperfections. These artificial intelligence capabilities, in conjunction with voice recognition, were also the reason Google Home was able to be launch rather quickly.

IoT developer tools and operating systems

Google has realized that IoT developers play an important part in the decision-making process for or against new IoT offerings. As a result, the company is trying to make it easy for hardware and software engineers to develop for the Internet of Things. Google is offering an IoT developer prototyping kit, in conjunction with Seeed Studio and BeagleBoard, so engineers can start developing their ideas more quickly. In late 2016, Google added another important tool for developers: Android Things, previously called Brillo. This platform allows professional, mass-market products to be built by developers without previous knowledge of embedded systems design. Android Things should not be confused with Fuchsia, Google’s new lightweight opensource operating system for IoT. Fuchsia is deliberately not based on Linux but on a new kernel called Magenta in an attempt to address a wider range of IoT devices.

Standards

Unlike its competitors, Google has no presence on IoT standardization bodies. Instead, the company is trying to establish its own standards. Its flagship project is the Thread Alliance which already has hundreds of members, including Samsung SmartThings and LG SmartThinQ. Thread aims to create a better way to connect and control products in the home. It’s a network architecture designed for low-power, low-speed, small-packet communications compliant with IEEE 802.15.4 in a Mesh topology. Weave, another standardization project, is described as a communications platform for IoT devices. It provides definitions for common devices and device traits to ensure consistency across products of the same type. It’s early days for Weave and it currently only supports four home device types currently: Light bulbs, mains outlets, televisions, and wall switches.

Other IoT applications and adjacent technologies

1. Waymo self-driving car platform: Google has teamed up with Fiat Chrysler and is currently testing self-driving minivans in four US states. Resulting data is stored and handled on Google Cloud, making use of its artificial intelligence capabilities, such as providing algorithms for autonomous driving.

2. Daydream Virtual Reality: Google has its own environment for virtual reality applications. It has yet to catch mass market success but it already allows NBA basketball games to be viewed remotely, live and in 360° using the Daydream headset.

3. Eddystone Beacons: Google’s open source Bluetooth Low Energy beacons bring geolocation indoors at much higher resolutions than a typical GPS signal provides. Google’s pilot application, with the US city of Portland, Oregon, equips public transit stops and vehicles with beacons to provide travelers real-time transit status accurate to within half a second.

READ MORE ARTICLES