The Philips FlexCare Platinum Connected toothbrush is more than just a device for keeping your mouth clean – it’s an IoT machine.
Wireless sensors measure the location, pressure and scrubbing patterns of the 31,000 strokes-per-minute vibrating bristles. The data is transferred via Bluetooth to a mobile app that provides a three-dimensional post-brush analysis of coverage, recommending areas of the mouth that should be “touched up” or given extra attention. There’s an option to send a month-long history of brushing patterns to your dentist to keep them informed of your brushing habits.
And all of this data, along with many of Philips’ other connected device efforts, run out of Amazon Web Services’ cloud. It’s a new era of Intenet of Things-enabled machines, and Philips wants to be on the cutting edge of offering its consumer and business customers access to more data, which they hope will help keep patients more healthy and the machines running more smoothly.
+MORE AT NETWORK WORLD: How a giant like GE found a home in the cloud| Why IoT scares “the living crap” out of this security guru +
Philips is tapping into the burgeoning IoT market, which research firm IHS says comprised 15.4 billion connected devices in 2015 and predicts will double by 2020. IDC estimates IoT was a $737 billion market last year and will undergo a 15.6% compound annual growth rate to reach $1.29 trillion by 2020.
Dale Wiggins is vice president and general manager of the Philips HealthSuite Digital Platform and is responsible for the company’s health suite digital platform, which developed the FlexCare toothbrush’s connected functionality. “It feeds into our goal of improving overall health care,” he says. “By connecting our devices and modalities in the hospital or consumer environment, it provides more data that can be used to benefit our customers.”
Another major effort has been to enable connectivity in the company’s imaging devices, specifically Magnetic Resonance Imaging (MRI) machines. The key advantage in this use case is reduced maintenance costs.
Traditional MRI machines that are not “connected” record their vital statistics – how efficiently the machine is running, status of various components and reporting of any errors. In the past it has required a trained technician to tap into the machine, read the log files, then schedule a maintenance repair based on parts that need to be ordered.
With connected machines, those log files can be securely uploaded into Philips’ IoT cloud, which is housed in AWS. The machine, using backup computing capacity in the cloud, can now self-analyze how it’s performing, what the levels of various fluids are, and cycle times for moving parts. Philips collects this data and aggregates it across many machines from all around the country. When one machine shows an issue, Philips can look for warning signs that may have predicted it, then flag those as alarms in other machines. Instead of specialized technicians making hospital calls for checkups, they can now respond only when they’re actually needed – making their work more efficient, and less expensive for customers.
“What this is really doing is pulling information much quicker and making it actionable through analytics and algorithms,” Wiggins says.
IoT from the cloud
The back-end system to enable the connected functionality of the MRI machine runs largely out of the AWS cloud. The log data is collected as it normally would be, but newer machines and retrofitted older ones have a set of receivers that can securely transfer the data into AWS’s cloud.
Using a series of Lambda functions (that’s AWS’s serverless computing platform), the data is collected then organized, processed and stored. Philips uses predictive algorithms and data analysis tools to monitor activity, identify trends and report abnormal behavior.
Dale Wiggins, VP and General Manager of the Philips HealthSuite Digital Platform
Philips uses a combination of AWS services, including the company’s IoT platform, along with about 10,000 Elastic Compute Cloud (EC2) and various database instances, Wiggins says. Tools like Amazon’s CloudWatch (for alerting) amd Cloud Formation (for automatic scheduling and execution of tasks) are used. Philips has moved 19 petabytes of medical imaging data, partially using Amazon’s Snowballs, from hardware-based systems into the cloud. Images are stored in databases, Amazon’s Simple Storage Service (its scale-out block storage platform) and Glacier, its long-term “cold” storage service. All data is encrypted by Philips, which holds the keys.
What’s the hardest part about the whole process?
“It’s a highly competitive market,” Wiggins said. “We have to stay on top of our game, which means educating our workers, recruiting new workers and rewarding them appropriately.”
He added: “Philips as a company has decided this is going to be a significant skill area for us. We’re transforming from mainly a device-focused business to a health technology company focused on the health continuum of care and service. Connectivity is a big part of that, which is requiring us to transform as we go through this process.”