All Things Location

 View Only
  • 1.  GeoBuiz 2020

    Posted 01-17-2020 17:23

    This past week, I attended the GeoBuiz conference, a two-day event of primarily C-level execs from private and government organizations. The speakers and agenda can be found here. I represented Syncsort on a panel for Location and BI Platforms.


    There were a few take-aways from the event, in particular around a convergence of disparate geospatial technology sectors.


    For example, building information models, precision measurement (think LIDAR), autonomous vehicles and 5G are reshaping geospatial tech. It's almost trivial to say that "data" is driving the convergence of these technologies. Rather it's the capture, transmission and management that shapes how these applications are being defined. 5G is fast becoming a driver because of the ability to transmit data faster; LIDAR is replacing traditional imagery for measurement and visualization. BIM is now mandated by many countries for better environmental assessment. Autonomous vehicles, at least for limited, urban areas and low-speed use, cases require both precision mapping and vehicle-to-vehicle communication. "Mapping on the fly" is possible when you consider that LIDAR could be embedded into every AV.


    Machine learning for image feature extraction still looks like it's just in its infancy. But launching new satellites is the dream of some of the VCs at the meeting. More than one company was there looking for seed money. For me, there are still "too many pixels."


    I also participated in a meeting with the U.S. Geological Survey (USGS) and the director of the National Geospatial Program and got an update of their Data Platform that will host nationwide, LIDAR, hydrographic, elevation and other POI data (transportation, etc.). The invited attendees included Google, Oracle, PlanetIQ, Riegl, TomTom and about 20 others. The USGS was interested in determining how often and how much USGS data was being downloaded and used by geospatial vendors. The objective for them was to determine future investments in data and the distribution of government-collected data as well as the commercial solutions that are being developed with USGS data as a foundation. The recommendations of the group to the USGS was to make certain that all data was available as a cloud-deliverable model (APIs?). There is also a recommendation to have more of these types of meetings and the USGS is looking for those opportunities. The USGS is becoming more deliberate in their approach with vendors. This session was the first I've ever attended where the USGS was actually soliciting feedback from the vendor community. They have indicated that they will host more of these sessions and I have the name of the director for the NGP.


    What struck me about this event is that it was still narrow in focus, more a meeting of the traditional GIS vendors (Trimble, Riegl, Hexagon, Autodesk, Esri) than an extension of the industry to include some of the new companies in mobile marketing and location data, though last year I believe that Factual and Foursquare where in attendance. Microsoft, Oracle, Google were represented at the meeting but not companies like Zillow, Facebook, Trulia, Uber, etc.


    There was value in attending this meeting and it's important to stay connected to this segment of the geospatial industry. Yet, there is certainly an opportunity to engage other segments of the broader LI sector to involve companies like Tableau, Alteryx, Salesforce that are embedding LI capabilities.

    In summary, the conclusion I came to was that the geospatial domains show more inclination of coalescing. Raster data (satellite, drone, etc.) are now leveraging machine learning to extract the vector data needed for geospatial analytics, because on its own, raster data was often too hard to utilize for the average users. You had to become well-versed in image processing and learn a very complicated piece of software to extract data for mapping. Now, the technology is such that building footprints, changes to the landscape and other features are more easily extracted. But to be sure, there will be more and more geospatial data to process as more sensors are deployed, more satellites launched and more drones are flying. 

    Location intelligence is still on a wild ride.

    Joe Francica
    Boulder CO