Drone Data Management: A Conversation With David Tran

0 328

Commercial drones can provide valuable data to enterprise customers – but dealing with that data is key to serving customers.  The huge quantity of data that drones gather is often not utilized to it’s full potential.  With thousands of images to sift through, service providers often struggle to deliver pertinent, actionable information to the correct person at the client team.

That’s where a drone data management platform comes in.  We talked to David Tran, co-founder and CEO of Optelos, one of the most innovative and integrated platforms available, to get his thoughts on using technology to manage drone data effectively and to explain some of the tools available.  Tran is a veteran of the wireless industry, which introduced him to drones through cell tower inspection work.  Tran and his co-founder developed their product based on deep experience in enterprise operations and in helping customers manage and improve their systems projects – experience that has helped them understand just what enterprise customers need.

Named one of CIO Applications Top 25 Drone Companies of 2017, Optelos combines visual data management tools with AI technology to deliver a feature-rich platform.  “Our drone data management and AI Analytics platform turns unstructured data into answers and delivers that data in useful ways,” says the company.

DL: A lot of drone companies now are using some form of AI technology in their offering.  Why is AI so important for managing the large quantities of data that drones produce?

[DT]   We’ve seen firsthand how the massive volume of unstructured data can quickly overwhelm drone service providers and Enterprises.  Our customers have told us they spend on average 3-4 x times longer processing and analyzing data than collecting the data. 

Why is that?   Many companies are still manually processing, analyzing, and tagging imagery one at a time.  As you can imagine this process is labor intensive, time consuming, and not scalable when dealing with millions of datasets. 

AI machine learning, when properly implemented within the data management workflow, can significantly streamline this work and reduce the time to deliver insights.   This means companies can scale their operations and data collection while reducing the overall time required to process and analyze that data.

DL: Optelos uses visual data management tools.  Can you explain what visual data management is, and why it matters to customers?

[DT]   Drone data is complex and largely unstructured.  The issue with most data management approaches is they only focus on storing and simply reflecting the captured data imagery back, leaving it up to the user to connect the dots in order to extract meaningful results from the data.  But we know that customers want – and frankly expect – much more from data management.  They are looking for answers.  This means automatically connecting the dots into a rich taxonomy, providing an immersive visual roadmap that delivers precise value added context.  

Our approach at Optelos is quite different. From the moment data touches our platform, we extract all the rich metadata, correlate that data, and deliver that through our Visualizer and FlyView™ presentation layers.  This means instead of presenting a basic list of images that requires the end user to piece together the story, we provide comprehensive visually correlated data that links together images, videos, maps, 2D Orthomosaics, and 3D models.

Of course, all that is combined with the AI, statistics, and dashboard to create a complete visual workflow experience for the user literally at a touch of a button.

DL: I understand the advantages, but what’s the real ROI?  What’s the time savings for the drone operator?  For the customer?

[DT]  We’ve done extensive studies with our customers and their stakeholders, and found that they reduce on average their mean time to delivery by 60% compared to the current manual workflow.  This is huge since that equates to many hours saved per job that directly impacts their bottom line.

DL: How can customers customize AI platforms?  What if they have customers in totally different industries?

[DT]  We approach AI differently than most.  First, we understand that AI works best when it’s coupled to the customer workflow rather than as disjointed steps.  Second, each industry is unique in terms of the analysis required.  For example, identifying bad cells in a solar panel array is different than detecting a flashed insulator on a utility pole.  So we believe tuning the AI for specific industry requirements yields increased accuracy.   And third, often times customers require customized object classification categories and a presentation layer tailored to their business needs.   

We recognized this need early on and implemented a highly flexible and modular platform architecture designed from the ground up to allow for customization.  This means we can deliver a unique combination of workflow integration, industry focused AI algorithms, and customer specific AI classification algorithms all housed within a single platform.

DL: Most platforms now offer cloud storage and processing.  What are the potential drawbacks with cloud-based data processing systems?

[DT]   While cloud based systems provide extremely secure, scalable, and inexpensive solution, it does have a few limitations such as integration with existing Enterprise back office systems.  In addition, some enterprises require that the data and platform reside within their data center for compliance reasons.

Fortunately, our Enterprise background allowed us to understand these needs early on and specifically design our platform so that we can deploy behind the customer’s firewall just as easily as on the cloud.  It’s one of the unique, customizable features that Optelos offers.

DL: How has drone data management evolved over the last few years?

[DT]  You may have heard me talk about the fact that customers don’t want data, they want actionable answers.   Customer’s expectations of data management has shifted away from storage and rudimentary access to the data.   Data management is evolving to mean this coupling of workflows, analytics, and delivery that is custom built for the customer with any number of industry use cases.

DL: Where is this part of the industry heading next?

[DT]  As you can imagine, the volume of data will continue to increase.  Data analytics and data management will continue to get smarter.   We believe businesses will increasingly adopt data analytics tools in order to leverage the massive quantities of data available to them in order to grow and remain competitive.  Simon Thomas, VP and leader of the IBM Watson AI team, estimates that analytics focused companies see a third more revenue and 12x times better performance than those who don’t.   

In short, we see an ever increasing customer demand and expectation to leverage the power of smart analytics. This in turn will transform data insight and integrate analytics into everyday business processes.

Going to AUVSI’s Xponential 2018 in Denver next week?  Don’t miss David Tran participating in The Power of Data Across the Unmanned Systems Ecosystem session, Room: 403/404, Tuesday, May 01, 2018: 1:00 PM – 2:00 PM, or meet the Optelos team at Booth 1826.


قالب وردپرس

You might also like More from author

Leave A Reply

Your email address will not be published.