By Megan Bozman

The world of IoT might be fraught with TLAs (Three Letter Acronyms) and a wide variety of use cases, but I’m intrigued by a central, recurring theme: IoT ‘things’ must be not only connected, but the data must be analyzed in order to see value from connectivity.

Automotive is one industry rapidly joining the IoT as cars become more connected. When reading “connected car,” many people may think of Wi-Fi hotspots or Bluetooth integration enabling music streaming and hands-free phone calls. But connectivity in the automotive industry has wide-reaching implications; far beyond entertainment and convenience.

In a recent blog post about Hortonworks’ partnership with DataRPM, Grant Bodley, general manager, global manufacturing at Hortonworks, wrote, “With the exponential growth of connected vehicles comes the need for Connected Data Platforms, data science, and machine learning algorithms.”

Mr. Bodley points out the need to “distinguish ‘signal versus noise’ amidst billions of sensors.” This partnership therefore aims to “help automotive manufacturers gain the valuable insights needed to achieve measurable business outcomes with greater speed and machine efficiency.”

Further, Mr. Bodley asserts that Hortonworks’ Connected Data Platforms are ideally suited to managing the Connected Vehicle Data Pipeline and the issues around the speed and scale of data generated by connected cars. Automakers can analyze this data in real time by leveraging the Platform.

At the recent TU Automotive Detroit conference & exhibition, Jay Jobanputra, director of sales at DataRPM provided a demo of these capabilities, embedded below.

Mr. Jobanputra showed a dashboard capable of analyzing the data streaming live from a connected car’s numerous sensors. Data points are collected and broken down based on criticality. Through automating data science, insights possible include predictive maintenance and the chances of vehicle failure. Users can select timeframes for viewing, and are also able to edit the code behind the scenes to fine-tune algorithms.

Additionally Joe Niemiec, systems architect at Hortonworks, showed a demo of Hortonworks for Apachi nifi and embeddable hardware featuring the Qualcomm Connected Car Reference Platform. This demo is also embedded below.

The Qualcomm Connected Car Reference Platform integrates the multiple wireless connections found in modern automobiles into a centralized architecture. The demonstration simulated bi-directional data communication between an on-vehicle platform and the cloud.

Data from the car can be sent to a cloud endpoint via two different flows; via LTE and Wi-Fi, when available, in order to reduce costs. Real-time info such as GPS location-based data can be sent over LTE, whereas sensor data related to diagnostics could be batched and sent via Wi-Fi. Automakers are able to bring the data back to data centers to analyze.

Effectively gathering data, transmitting it to the cloud, and subsequently analyzing it are clearly all key to the success of IoT within the automotive industry. Hortonworks has shown interesting applications with DataRPM and Qualcomm and I look forward to reading more of the success of these new applications.

Share on Facebook0Tweet about this on TwitterShare on Google+0Share on LinkedIn0