Waypoint

Up and Under — On Using Drones for Bridge Inspection

When it comes to inspecting bridges, drones (or UAVs/UAS) are often heralded as the next big thing, a truly disruptive technology capable of giving engineers eyes in the hardest to reach places, without the need for expensive access vehicles or potentially dangerous rigging (or even ladders).

But are unmanned aerial vehicles actually practical to use and are their supposed benefits achievable today? Waypoint caught up with two of the United States’ most experienced experts to find out.

Barritt Lovelace, regional manager at Collins Engineers and Jennifer Zink, the State Bridge Inspection Engineer with the Minnesota Department of Transport (MnDOT) Bridge Office, are currently running a multiphase UAV research project. Its goals: evaluate the potential of drones to cut costs and improve safety when inspecting bridges, and create best practice guidelines that detail when and how to best employ UAVs.

This research has sparked interest across the U.S., including from other states’ transportation departments, as well as extensive media exposure.

Hello. Why don’t you start by giving us a quick overview of MnDOT’s drone research?

Zink: Sure. We have an ongoing study that we’re running in conjunction with Collins Engineers, which is looking at how UAVs can be used to provide better data and make inspections safer. We finished phase one last summer [after which MnDOT published a downloadable research report], inspecting four bridges across the state, and we’re currently working on phase two.

We’re looking at how UAVs can be used to provide better data and make inspections safer

What type of drone are you using?

Zink: For phase one we used an Aeryon SkyRanger. This is a great UAV, but we identified a couple of shortcomings as it applies to bridge inspection, such as not having the ability to look up and not being able to fly underneath bridges where there is low or no GPS coverage.

Phase 1 video: an overview of MnDOT’s research (drone shown: Aeryon SkyRanger)

Then we saw the albris and contacted senseFly to see about trying this. We really liked that it has ultrasonic sensors and additional cameras, plus it can look directly up and it can operated without GPS, so we are able to fly it under a bridge. We’re assessing this drone during phase two of our program.

The senseFly albris, used for phase 2 of MnDOT’s research.

Are other states carrying also similar drone research?

Zink: From what I have learned, Caltrans did look into drones for bridge inspections, maybe five years ago, but they had problems back then with drones that didn’t fly. Otherwise we’re pretty much pioneers here, nationwide, in the U.S. We’ve had lots of calls from other states who are looking at this, but we’re the only state to have done a project on this so far.

For a little wider context, could you explain what kind of bridge inspection techniques would traditionally be used? In other words, what type of equipment might drones replace?

Zink: Traditionally the equipment choice you had was either to use an AWP, an aerial work platform, or to go in via rope access. This decision would depend on the type of bridge and your ability to close lanes of traffic.

On state highways, we’ll typically use under-bridge inspection vehicles like snoopers, but if a bridge is small enough it might only require ladder or rope access. Some are very low so we can’t even get a snooper underneath.

Bridge_inspection_rope_access_MnDOT
Getting a closer look at a Minnesotan bridge via rope access. (Photo: MnDOT.)
Snooper_truck_AWP_MnDOT_UAV
One of MnDOT’s ‘snooper’ access vehicles in action. (Photo: MnDOT.)

What is the relationship between MnDOT and Collins Engineers?

Zink: Bridge inspections, both fracture critical and special, are traditionally handled by MnDOT. Each District and each County has their own inspectors to conduct routine bridge inspections. But sometimes, if an agency has a small bridge roster, they might use consultants such as Collins. The firm’s work for us though was traditionally underwater.

Lovelace: We carry out bridge inspections nationwide. As Jennifer said, MnDOT tends to do many of their own, so we’re called in to do more specialty work such as rope access or underwater inspections.

Tell us more about your research, particularly the current second phase. What are your focus areas and goals?

Zink: While phase 1 was about evaluating the safety and effectiveness of UAV technology to produce an initial research report, with phase 2 we’re looking to develop a best practices document, including a full decision tree. This will look to answer when a drone might be more suitable for inspections than a snooper truck, as using a snooper needs lots of coordination, such as traffic control, and it puts inspectors at risk.

Using a snooper needs lots of coordination, such as traffic control, and it puts inspectors at risk

We want to try using the albris to inspect different types of bridges, including bridges that are confined spaces. We’re inspecting culverts, box girders, and doing deck delamination assessments using the drone’s thermal camera.

The ultimate goal at the conclusion of this research is to identify bridges where UAS inspection could provide the close detail necessary for a thorough inspection. Currently at MnDOT, we are attempting to initiate a policy for bridges that meet certain condition criteria (i.e. poor) that will require close-up inspection every five years. The access requirements could potentially be very costly to meet this policy, especially for local agencies. We anticipate that inspection by UAS can alleviate a lot of these costs.

We anticipate that inspection by UAS can alleviate a lot of these costs

The idea would be to implement a state-wide UAS contract to obtain close-up inspection data for all bridges meeting the criteria of this policy on a four-year cycle, for both state-owned and locally-owned bridges. The details of this in regards to funding and state aid involvement will be addressed at the conclusion of our Phase 2 study.

Phase 2 video: using the senseFly albris under Blatnik Bridge

How exactly are you using the drone in your study?

Barritt: We’re really approaching it like a regular inspection. If we are in a situation where we’re looking at a certain spall in a certain area, we’re not just relying on taking that photo back to the office and identifying the deficiency and location later; we’re going to note the photo number, the location and direction, on site, at the time. The National Bridge Inspection Standard (NBIS) requires that a properly trained and certified inspector be present and lead the inspection in order to ensure the safety of the travelling public. Having an inspector simply review drone photos at a later time would not meet that requirement.

One thing we did that worked really well was we projected the laptop running our eMotion ground station software to a second monitor, so our inspector had one screen, and was also controlling the drone’s camera, while our pilot flew the drone off of the laptop monitor. Inspection is about doing it live, being involved, rather than later in the office. The more interactive, the better quality the inspection.

The more interactive, the better quality the inspection

And in your research so far, how are drones faring?

Zink: I don’t see any limitations at a drone level right now. It works well for what we need and the images correlate well with the inspection reports that we’re comparing them too.

Several of the UAS’ imaging devices were, or are, being tested, including its still image, video and infrared cameras. Then in the field we’re collecting various data, such as still images, video, infrared images, site maps and 3D models of bridge elements.

The results so far have shown that the inspection detail that UAS provide effectively replicates some of the detail learned through the use of snoopers, without the traffic control requirements, and at significantly lower cost in terms of equipment and traffic control needs. UAS can provide both infrared and 3D modeling detail of bridges, effectively identify concrete delamination, gather topographic mapping detail and efficiently map riverbank conditions, upstream and downstream, from the bridge site.

The inspection detail that UAS provide effectively replicates some of the detail learned through the use of snoopers, without the traffic control requirements, and at significantly lower cost in terms of equipment and traffic control needs

When you talk about lower costs, did you perform any detailed calculations you could share?

Even though our study is still in-progress, I did complete some preliminary cost estimates of UAS versus traditional inspection access methods on an inspection of the largest bridge in our state.

Based on traditional methods of inspection, this bridge would typically utilise three snooper inspection vehicles and require a total of eight inspection days. This equates to a minimum cost for that inspection using conventional equipment of approximately $59,000. That doesn’t include equipment mobilisation and travel expenses.

The cost of a UAS contract to inspect the exact same approach spans of our sample bridge would be around $20,000 with only five days onsite as per our consultant-obtained quote. This is a potential cost saving of 66%.

As I mentioned, at MnDOT we are attempting to initiate a policy for bridges that meet certain condition criteria (i.e. poor) that will require close-up inspection at an interval of every five years. The access requirements could potentially be very costly to meet this policy, especially for local agencies. We anticipate that inspection by UAS can alleviate a lot of these costs.

The cost of a UAS contract to inspect the exact same approach spans of our sample bridge would be around $20,000, with only five days onsite. This is a potential cost saving of 66%

When your study ends, what’s next? How will you plan to put your findings into action?

Zink: Our goal in the future, after our study ends, is to identify bridges in Minnesota that would benefit from UAV inspection, then set-up a statewide contract on a regular cycle – i.e. every five years – to do drone inspections for us. We don’t plan to buy drones ourselves, but rather get a contract and have consultants do that.

Thanks so much for speaking with us today.

You’re welcome!

Watch a WCCO – CBS Minnesota clip about MnDOT’s UAV research:

Read MnDOT’s project summary

Learn more about the albris inspection drone

5 Comments

  1. Danny

    18.07.2016 - 22:24
    Reply

    I believe that the CALTRANS study was in 2008. After that, various other states have conducted (and continue to conduct) significant studies related to unmanned aerial systems (UAS) for bridge inspections. For example:
    (1) Proof-of-concept study conducted by Florida Institute of Technology, funded by the Florida Department of Transportation (2013 – 2015): http://www.dot.state.fl.us/research-center/Completed_Proj/Summary_MNT/FDOT-BDV28-977-02-rpt.pdf
    (2) http://unmanned-aerial.com/connecticut-dot-testing-uavs-for-bridge-inspections/
    (3) http://www.detroitnews.com/story/news/local/michigan/2016/03/28/michigan-drones-bridge-inspections/82356034/

  2. Darrin Lane, P.E.

    25.01.2017 - 17:59
    Reply

    UAS technology and capability is developing at a pace that is so rapid, studies performed even two years ago are going to significantly understate the value and applicability.

  3. Rich Hovde, PE

    19.04.2017 - 09:06
    Reply

    I remember telling my boss Hugh Favero about this new digital camera that doesn’t use film. He said it would never work, it will get wet and fry, it’s too expensive for us roughnecks, and “hell no, I’m not buying it”. That was 1990 when we had one telephone between 10 guys. Now it is 2017 – the cameras on these drones are wicked incredible multitasking devices; its exciting times for those of us tinkerers that like fiddlin’ with stuff to make it faster, stronger, lighter. We have several drones in our fleet and going full tilt production next month God-willing and FAA/ATT allowing!

Leave a Reply

Your email address will not be published. Required fields are marked *