Internet of Things

Make Devices Better

The Goal

  • Healthcare: Modern medical device and EMR data can help to predict the incidence of serious conditions like sepsis and coordinate the appropriate medical response. Similarly, models can monitor caregiver behavior and determine adherence to clinical pathway guidelines.
  • Transportation: Automobiles, airplanes, and trains can generate gigabytes of data per hour when operating. Hidden in those data are opportunities for reducing fuel costs, improving navigation, and scheduling maintenance.
  • Manufacturing: Embedded sensors monitor for signs of quality and reliability issues. Machine learning models built on these sensor streams can predict the likely occurrence of a defect or equipment failure and automatically take steps towards remediation.

 

"Managing the massive data flow from connected devices using traditional batch processing just isn't feasible. IoT has been a major driver behind our real-time scoring engine architecture and our support for streaming APIs."
Josh Lewis, VP of Product
Alpine Data

The Impact

  • Healthcare: Decrease likelihood of complications and readmissions. Raise compliance with evidence-driven best practices.
  • Transportation: Reduce fuel costs and unexpected maintenance. Improve fleet utilization rate and efficiency.
  • Manufacturing: Shorten periods of downtime and prevent unnecessary rework. Close the gap between order and delivery.

The process of building models is actually quite familiar in IoT use cases. You’re working with historical data and there’s usually a fair amount of ETL involved. What’s magical is the impact the models can have when deployed — directly manipulating machines on a factory floor or sending a notification to a nurse.

Emilie de Longueau. Lead Data Scientist
Alpine Data