How Google Cloud IoT Core helps Smart Parking process vast volumes of data

parking chesky w
Image: iStock/Chesky_W

The parking optimisation company turned to Google to bring its third generation IoT platform to life

Share

Smart Parking is using Google's new Cloud IoT Core platform to reinvent the parking experience through use of sensors and advanced data analytics.

The company helps customers including Transport for London, the City of Westminster and Hilton Hotels optimise parking at their sites through vehicle detection sensors that monitor the occupancy of spaces.

These devices generate a vast volume data that can provide increasingly sophisticated insights to businesses and drivers. To cope with the growing quantity of sensors and data and take advantage of the latest developments in data analytics, Smart Parking needed a management platform, and turned to Google's Cloud IoT Core.

"Now we have a platform that is totally secure," says John Heard, Smart Parking's CTO. "It scales from zero to a billion devices on demand and we can place it in any region to regionalise the data that's available."

The IoT market

Cloud IoT Core was developed by Google to help businesses in sectors such as utilities, transport and logistics, oil and gas and manufacturing manage their expanding fleets of smart devices and the volume of IoT data they produce, securely, at scale and from a central location.

This data can then be queried using Google’s own range of advanced analytics services.

Read next: Best cloud IoT platforms

Smart Parking was one of the early adopters of the platform, which was launched to the public in September.

The result has made major efficiency savings for the company. "It would typically take two to three weeks of configuration and operational acceptance testing," Heard says. "It's now taking us about three to four days.

"We can activate large numbers of sensors and devices very, very quickly and we can pre-stage them in our assembly warehouses before we put them on site, which previously we had to do that while we were on site."

Smart Parking can also leverage Google’s data visualisation and reporting tools such as Data Studio and Data Prep, and integrate new machine learning services, like TensorFlow, as they’re rolled out by Google. Heard plans to use Google’s platform to provide customer’s with real-time parking guidance tailored to their individual need and location.

"The opportunity is now opening up to us to achieve our mission of reinventing the parking experience, and a big part of that is using machine learning to give better guidance into the consumer's app," says Heard.

"Those are the sort of things that are changing the whole mentality from trying to find a parking space to being guided to it automatically and conveniently in a way that will be quite transformational. And these are the aspects that we can do that we couldn't do with the old system."

Why Google Cloud IoT Core?

Smart Parking managed the earlier iteration of its platform through a combination of Windows and Linux-based servers running on the Amazon (AWS) cloud platform. Last November, the company decided to build a third generation of its platform.

"It came to a point where we recognised that if we kept just evolving the second-generation platform we weren't going to enable ourselves to exploit and utilise the benefits of what I call 'cloud nation computing', where it's completely on demand and it scales up and down dynamically," says Heard.

“What we realised was that we needed to have an information computing platform that would enable us to scale very dynamically, very much in real-time and also open up more opportunities for ourselves.”

AWS and Azure, Google's arch-rivals in cloud computing, both offer their own IoT management platforms, but Google stood out due to its reputation for innovation in data science.

"We looked around at a variety of the classical ones, Microsoft, Amazon and Google, and the characteristics that very much drove my decision-making were that we wanted to push forward into the advanced capabilities of cloud-native computing.

"It was more advanced, it was more complete, and it had a comprehensive frame that would take us forward into machine learning that was quite natural compared to other providers at this present time.”

Read next: AWS vs Azure vs Google: What's the best cloud platform for the enterprise?

It took just a week for the company to integrate its platform and smart devices into Cloud IoT Core, launching in September.

The Smart Parking team had spent 10 years developing the previous generation of its platform. The latest generation was ready in a year. Heard attributes much of their speed to Google’s dynamic development environment, which allows them to constantly add new features.

"The Cloud IoT Core has been an acceleration that helped us to eliminate a whole variety of issues that we previously had to develop and manage for our infrastructure," he says. "Now I can focus a team on other things."

Find your next job with computerworld UK jobs