Skyward recently hosted a webinar with the Ohio Department of Transportation to discuss major roadblocks to building a drone program. They shared about their cutting-edge use cases and what it takes to operate an enterprise-scale drone program. They even discussed how they calculate and obtain ROI for drones.
Our audience submitted dozens of questions for our panelists. Because we had so much great content to cover, we were only able to answer a few of these live on the webinar, but we promised to follow up later on some of the top questions. Our panelists from Ohio DOT graciously provided the answers below.
How big is Ohio DOT’s program? How do they handle mission authorizations?
David Gallagher (Flight Operations Manager, Ohio DOT): Our program is one of the largest DOT programs in the U.S. This year so far we have flown 1,806 flights and 483 projects. That’s a lot of risk we assume every single day. Some days I have 6 or 7 risk assessments to approve. That review includes looking at airspace, weather, and obstructions line power lines or congestion of people. I try to give or seek the best advice before approving an operation.
We use the Skyward platform to do our risk assessments, our flight scheduling, and our flight planning. For the approval process, Rich Fox and I are responsible for low and medium risks. However, if it is a high risk operation, that goes to our Director of Ohio UAS Center for approval.
What mapping tools and software do you use?
Helen McCreary (UAS Program Analyst, Ohio DOT): It depends on which use case we’re flying. If it’s for magnetometry survey, we use UgCS because it has a great terrain following function. For surveys, our team uses the Phantom 4 RTKs and controllers with built-in tablets.
We are constantly working to improve and expand our capabilities, so we try out a lot of tools and software.
For post processing, we use Pix4D and ContextCapture for photogrammetry. I currently process the magnetometry data using QGIS and Surfer, but we are looking at Oasis Montaj. Our surveyors view the point clouds in Microstation, and some of them use TopoDOT to extract linework from our point clouds semi-autonomously. We’re going to try a demo of Pix4Dsurvey soon, also for linework.
What aircraft do you fly in your drone fleet?
David Gallagher: The Ohio UAS Center operates DJI and Skydio products. The DJI aircraft in service are:
- Eight – DJI Phantom 4 RTK
- Four – Skydio 2s (with more on the way)
- Three – DJI Matrice 210 v1 and v2
- Two – DJI Inspire 2
- One – DJI Phantom 4 Multispectral
- One – DJI Matrice 600
There are additional aircraft in the inventory, but these are the frontline aircraft.
Who does your drone maintenance at Ohio DOT? Does Ohio DOT maintain a written preventative maintenance policy?
Rich Fox (Airspace Manager, Ohio DOT): We are blessed to have our lead pilot, Daniel Kammer. He is a certified FAA A&P mechanic. There is a huge void in the industry when it comes to maintenance. The guidelines are slim to non-existent. So what Daniel has created is an entire chapter for us on aircraft maintenance and testing. Using the FAR 43, Appendix D, Daniel has created an annual and a 50-hour inspection checklist that he uses to manage our fleet. The minor things can be taken care of in the field by our PIC, but anything other than that, we bring it in-house for Daniel to resolve.
How much time does it take to train reliable pilots? Is the process complicated?
David Gallagher: Currently, we are averaging approximately 30 days to train a new pilot, which does not include Part 107 test prep. We provide material to study for the 107 test and the individual works at their pace until they are ready to take the test.
Once they have taken and passed the test we start their ODOT training. We start with online training which covers our manuals, procedures, software, safety management and takes about 10-15 days. The rest of the time is hands on flight training covering basic airmanship, teaching best practices, emergency procedures, and safety compliance.
After training, the new pilot accompanies one of our experienced pilots on actual missions to observe, then fly an actual mission under the supervision of a certified pilot.
How can companies or government entities get involved in Ohio DOT’s drone initiatives?
Rich Fox: Companies can visit our website at UAS.Ohio.gov to follow our initiatives and submit inquiries regarding unmanned activities.
Is Ohio DOT working with USDOT/FAA to develop targeted programs and facilities? Is the FAA invested in Ohio’s success for the long term?
Rich Fox: The Ohio UAS Center works with the USDOT & FAA on multiple fronts and has been instrumental in evaluating and approving our efforts for unmanned aircraft operations. With their support, under an approved COA, we are working with the Air Force Research Lab conducting BVLOS flights up to 10,000ft in a 200sq mi operating area utilizing Ground Based Detect and Avoid (GBDAA) system. The state of Ohio has invested over $12 million in infrastructure to support unmanned and autonomous operations and works closely with the FAA to further the development of the regulatory process.
What data is actually reviewed when flying a drone inspection? Are you building something from the imagery or using the imagery for the inspection?
Helen McCreary: A lot of the inspection is done live, through the camera. Our bridge inspector pilots know what to look for, and know when something needs closer inspection. They also take photographs with the Skydio. They have an order in which they take the photos so they can organize them in folders later according to location on the bridge. It all feeds into the same documentation protocol they have for normal bridge inspections. It might be useful in the future for every bridge pier or I-beam to have a number posted on it, for easier photo location during a drone inspection.
My colleague Jamie and I are experimenting with creating 3D models of bridges, but so far the bridge inspectors have not shown much interest in this. The challenge lies in processing and displaying a model with high enough resolution to actually see the cracks. A model that detailed would be huge, and most computers couldn’t handle it. We’re continuing to work on this, though.
As of right now, they are manually identifying cracks and assessing damage, using the photographs. A few research projects have looked into automatic crack detection, but we have not done much with it yet.
Interested in learning more about the Ohio Department of Transportation’s drone program? Watch the on-demand webinar and download the slides.