Geotab Data Bootcamp November 2020 recap

Fleet and connected vehicle professionals joined data science experts for the first-ever fully online Geotab Data Bootcamp between November 17-19. This Bootcamp featured three days of instructor-led and hands-on workshops.

Geotab Data Bootcamp focused on the importance of using data analytics to drive growth and innovation through hands-on training in AI, data engineering and visualization, with an emphasis on telematics. Sessions were hosted by subject matter experts in the field.

Spots filled quickly, so for those who missed it, see the event recap below.

See also: Highlights from the virtual 2020 Fleet LatAm Conference

Day 1: Planning predictive maintenance pipelines

Mike Branch, Geotab’s VP of Data and Analytics, kicked off the Bootcamp with an exciting introduction to the Geotab Data ecosystem and the value of driving data with AI. 

Throughout the rest of the day, attendees had the opportunity to hear from Geotab data scientists about:

• Working with geospatial data in Python
• AI basics with a telematics twist
• Building a MyGeotab Add-In that uses big data

The day was capped off with one of the Bootcamp’s most popular sessions: Data Pipelines for Predictive Maintenance, led by Geotab Data Scientist, Daniel Lewis. This session showed attendees how to go from raw vehicle data to a working predictive maintenance pipeline utilizing BigQuery, Airflow, Dataflow and Python. It also highlights how Geotab incorporates machine learning models into our data preparation processes.

Day 2: The power of AI

Keynote Speaker Patricia Florissi, Technical Director, Office of the CTO, Google Cloud, set the stage for the second day of the Geotab Data Bootcamp. Patricia’s presentation focused on AI technology and how it has been interwoven in everything Google ideates, designs, develops and offers in the market.

Florissi also provided a behind-the-scenes look at Google’s AI advancements. She described the ins and outs of Google’s AI and the many ways Google has been using it to create innovative experiences at a planetary scale. She finished by talking about Google’s vision to democratize AI and enable everyone to differentiate with Google’s AI innovations.

Geotab’s Bob Bradley, Associate Vice President, Data Solutions and Data Scientist Brenda Nguyen, also walked attendees through techniques used to uncover outliers in fleets ranging from safety analyses to utilization. They also explained why finding these outliers provides a clear path to optimizing a fleet.

Additional sessions on Day 2 included topics such as:

• Building machine learning models
• Accelerating time-to-market for machine learning projects
• Real-time mapping

Day 3: Data visualization and extraction

On the final day of the Bootcamp, Javed Siddique, Senior Data Scientist at Geotab, started the day with his presentation on exploring and visualizing massive amounts of data for model development. His sessions introduced attendees to interactive data visualization techniques using the Facets tool. Facets is an interactive data visualization tool that helps data scientists at Geotab visualize large volumes of data and solve complex problems.

Siddique helped attendees set up the Facets tool and load large volumes of voltage data in a notebook. This allows them to create interactive visualizations and to interact with the data to discover hidden patterns.

The day concluded with sessions on:

• Intelligent zoning
• Merging telematics with point-of-sale data
• Data extraction from MyGeotab and population into Big Query

At the end of the Bootcamp, attendees received a well-earned certificate of completion.

See also: What is data visualization?

Conclusion

This year’s Geotab Data Bootcamp event was a huge success, thanks in part to all of the attendees. The success of this training program would not be possible without the participation of all who registered. We look forward to hosting our next Geotab Data Bootcamp scheduled in March 2021.

For more information on Geotab and intelligent data solutions for fleets, visit data.geotab.com.