Mavic 2 Enterprise Advanced Thermal Drone - An Ecologists View

It’s 20:30 and I have a dawn bat survey tomorrow (or is that tonight, I never quite know) with the alarm set at 01:50 – midsummer in the Highlands brings an early dawn. So apologies if this is little rushed, but if I don’t do it now, another week or three will pass.

Since going freelance some seven years ago now, one of the things I wanted to do was use technology to increase the effectiveness and robustness of the survey work I do. I was never a fan of coloured pencils on large scale maps, if I’m honest. Technology is most certainly not the answer to all woes and although I try, I would be lying if I said that all my tech investments were problem led rather than technology led. To date investments in field computing, bat detectors, IR night vision cameras, camtraps, commercial weather stations and a host of mostly useful software, have all been a relatively straightforward decision with a pretty quick ROI. However, this winter I was toiling with the investment in a thermal imaging device – a much more costly and more difficult to justify business case, in my opinion.

I should back up a bit and just point out that my ecological business has three main components; the standard protected species and habitat surveys for a range of clients large and small; provision of GIS services to other consultants and organisations; and aerial imaging and mapping with the use of UAVs (drones).

I have been doing the latter for a good few years now and undertake a variety of work focussing on ecological and environmental applications. I am not a drone services company as such, I am an ecologist who uses a drone as a tool to provide a broad range of services. This is relevant to my final choice of equipment.

So when the plan of investing in a thermal imaging device came to the fore, the two main criteria were, good resolution for terrestrial based bat survey work and capable of being used on a drone to enable additional services to be offered. To cut a long story short this narrows the search down dramatically to only a few and I started seriously looking at the Flir A65 series (taking into account the additional problems with respect to using laptops etc in the field), but this would mean that I would need a new drone to utilise the device to its maximum potential. There were drones with thermal imaging cameras available but utilising these on the ground for bat surveys would be cumbersome (as the drones were relatively large) and difficult and the thermal imaging cameras that were present could not be used disconnected from the drone. As I was coming to the conclusion that it would have to be the thermal imaging first and then the drone to carry it a year or two later, DJI (the dominant drone manufacturer) released information about a new drone – the Mavic 2 Enterprise Advanced (M2EA).

The M2EA is a small drone, easily fits in a shoebox when folded up. There was an earlier model with thermal imaging camera but the resolution was poor and would not have been of much use on the ground or in the air. The specs for the M2EA are available on the DJI website so I won’t repeat here, but the key features of the thermal imaging sensor is an Uncooled VOx Microbolometer with a resolution of 640×512 @30Hz and an accuracy of ±2℃. It records radiometric JPEGs but not radiometric video. The sensor is DJI made as they no longer use FLIR products but DJI have produced some amazing technology to date so I was not too worried. The decision basically came down to should I pay £5-6K for a quality thermal imaging device which I will also need to buy peripherals for and a new drone at around £12K to get it in the air… or pay £6K for the M2EA.

After a fair amount of discussions with sales reps, the M2EA did seem to tick all the boxes I needed (more or less) but was definitely not designed for my main use case – bat surveys on the ground. Slightly pushed by end of tax year implications I made the jump at the end of March and after a tortuous wait received the M2EA at the end of April, just in time for the bats.

So how does it perform for terrestrial bat surveys?

Positive points…

  • The drone has an HD optical camera alongside the thermal imager and records both simultaneously – this can be very useful in some cases.

  • I have hired a number of different FLIR systems over the years for specific jobs and in my opinion the thermal image is as good as any FLIR I’ve used at the same resolution.

  • As it is a drone it is simple to have multiple batteries and switch them out quickly, but the device can run for 2.5 hours off one battery so this is generally not an issue.

  • Storage wise, large capacity micro SD cards can be used and there is also an internal memory which can be used in an emergency.

  • Perhaps one of the best bits about the device is the controller. As it is a drone the controller provides a view of the sensors (you can switch between thermal, optical or split screen). This allows you to monitor what the camera is seeing and recording from the other side of the Site (assuming you are happy to leave your equipment unattended). In addition the controller enables you to move the camera by pointing to a location on the image. This enables repositioning of the field of view or scanning the area remotely. You can also switch from recording video to taking stills if required. The range of movement is excellent with almost 180 degree left and right and a significant movement up above the horizontal. It is slightly difficult if you need to point the camera up sharply, but not vertical as the movement of the camera is in relation to the horizontal of the ground not the drone as when inflight the drone is not going to be heavily angled. You can point the drone straight up and the camera will then go into a position lock mode pointing straight ahead relative to the drone which is useful for some inspection and survey work.

  • I have found it will pick bats up easily at up to 25-30m depending on conditions and now use it almost every survey as either a survey point (in certain circumstances) or as a backup for surveyors at key locations (roosts, potential roosts etc.)

Negative points

  • No radiometric video but to be honest this would be very data heavy and generally not needed for consultancy work.

  • It’s a bit more fiddly to setup compared to a standard hand held but it is possible to purchase brackets to allow the drone to be mounted on a tripod and in turn this could also be mounted on a taller pole if required (this is how I generally use it).

  • It has no significant protection against weather so you will not want to be using it when raining – although generally this avoided for the surveys anyway. It is also less robust than standard handheld units so you need to be careful not to drop it while stumbling around in the dark.

  • The standard setting has large flashing LEDs but these can be set to turn off when recording so is not too much of a problem.

  • The recording times out at just under 30mins but can be instantly restarted if storage space allows. Strangely the actual file length for one 29min recording gets split into three (two 13mins approx and then the rest as a third file - not sure why this is). It also beeps on the controller to tell you it has stopped so you are aware and can restart.

  • Interference with my Echo Meter Touch devices only occurs within about 0.3m so this causes no major issue if using detectors nearby, you just need a little separation.

  • There is an internal fan that comes on for periods of about 1 min every so often. It’s a little annoying but not really and is quieter than the standby noise of a Phantom 4 for example. It certainly isn’t loud enough to impact on surveying even if you are standing next to it.

  • There are a number of standard colour palettes that you can choose from but you have no control over the range of the colour palette. This is probably the most significant negative as far as I can tell. The imager will decide what it thinks is the best range to use and will change as the cameras field of view is changed. This may be improved in software updates which are generally relatively frequent on drones as they improve the experience.

  • The field of view is relatively narrow for bat surveys but in most cases I have not found this to be an issue. There is no spec on the actual FOV but through using FOV calculators it appears to be about 42°. If I get time I’ll try and work this out accurately in the field.

It is perhaps also worth pointing out that there is a good RGB camera on this device as well which for daylight operations is a great bonus. The device is sold with a ‘optical and digital zoom) but strangely the photos stored when a picture is taken do not have the zoomed in image. This has been a major issue with drone service providers who want to capture the zoomed in imagery, but as an ecologist this is not a major issue. This may be resolved as a software update but is likely to remain unchanged as may be a limit to push the high end customers to buy the bigger more expensive kit. The RGB sensor is also sold as a 48mp camera which is not really the case, it uses some DJI wizardry to essentially combine four 12mp images. It is still a good camera with a good zoom nonetheless.

As far as looking at the readiometric jpgs is concerned, you have to use DJIs (free) software which arguably is significantly poorer than the FLIR software, but I suspect that this will change over time as they improve the customer experience on this new equipment. 

M2EA Ecological Applications

I have put together a few short clips from recent surveys that may be of use to establish whether or not the M2EA is for you. My main message would be if you are wanting to invest in thermal imaging and also fly or have committed to fly drones as well then it really is a no brainer. However, if the use of drones are not on your list of services then I would probably stick with one of the higher end handheld devices as these are likely to offer slightly better performance purely on the thermal imaging side of things. It is worth pointing out that the M2EA is actually a very good little drone that could be used for a whole host of other applications both thermal and optical. To date I have used the system professionally for assessing seals at haulouts, monitoring deer, rabbit surveys and countless bat surveys. I am also in discussions with clients looking to undertake other bespoke wildlife applications.

Feel free to get in touch to discuss.

Combining Ecology, Drones and GIS

During 2017 Tracks Ecology provided a range of bespoke services to a client looking to create a new broadleaved woodland, located on the west coast of Scotland. Initial baseline ecological surveys combined with high resolution mapping from unmanned aerial vehicles (UAV/drones) Tracks Ecology produced a highly detailed and geospatially accurate habitat map. This allowed ecological constraints to be accurately taken into account during the woodland planning stage. In addition the accurate mapping of bracken enabled contractors to plan control measures accurately. The digital elevation models (DEM) obtained from the aerial mapping also provided a very useful 3D model to aid planning with the ability to assess slope and aspect with precision.

Tracks Ecology is working with a client in relation to a proposed woodland creation scheme on the west coast of Scotland. The site required ecological surveys to ensure compliance with wildlife legislation and provide guidance for the proposed woodland planting. This model presents the results of the habitat mapping to JNCC Phase 1 Mapping classifications and 5m contours.

Once the woodland had been planned and approved the ground works (including fencing, ploughing and mounding) were undertaken. To ensure that the works were undertaken as proposed and provide an accurate measure of distances and areas, a second aerial mapping mission was undertaken. The results from this second mission also provides a baseline for monitoring the future development of the woodland in years to come.

The professional input of Tracks Ecology at every step of my woodland creation project has been invaluable. From a detailed and thorough Phase 1 habitat survey, which included accurately measured areas of  bracken, gorse and deep peat, through to post ground preparation mapping, with fence lines and areas of mounding and ploughing clearly measured, Tracks Ecology provided an accurate and indisputable evidential baseline for woodland planning, contractors, grant applications and claims. Invaluable on many levels.
— Client

If you would like to discuss potential projects in relation to ecology, aerial imaging or GIS then please do not hesitate to get in touch for a chat.

3D Modelling Helps Cromarty Harbour Trust

As a CAA certified UAV operator, Tracks Ecology has assisted the Cromarty Harbour Trust to obtain a detailed record of the 300 year old harbour. During March 2017, Tracks Ecology undertook a series of UAV flights to obtain both nadir and oblique imagery of the harbour at Cromarty on the Black Isle near Inverness. The brief of the project was to record a detailed picture of the harbour as it stands for use as a record prior to future renovations/repairs and as an educational and tourist information tool. Over 580 images of the harbour was obtained during a low spring tide with the addition of some images of the harbours setting on the Cromarty Firth. The resulting orthomosaic provides a detailed photographic record and combined with the 3D model provide a valuable resource to the Trust and Cromarty residents. Tracks Ecology used the DJI Phantom 3 Professional platform for the flight, with photogrammetry processing and point cloud editing undertaken within Pix4D. 

If Tracks Ecology can assist with any aerial survey work from habitat mapping to 3D model creation, please do not hesitate to get in contact.

Model of Cromarty Harbour on Black Isle near Inverness. Challenges of modelling both complicated objects (boats, steel bridges, ropes etc.) and adjacent to water (including tidal changes during photography) made the process a significant challenge.

Hot bats on a cold night

A few clips from a dawn survey of a large maternity roost I have been working on in Fort Augustus. I have been keen to trial some thermal imagery to see how they perform with bats and this roost seemed like a good test. Working with the team at Red Current we identified the FLIR E60 as the best unit to try.

The bats are clearly visible, but were not glowing as hot as I expected. The odd one comes back out of the roost and is significantly hotter than the flying ones, so I assume that they are fairly well insulated. The temperature during the night dropped and at the time of this was surprisingly cold at around 4 degrees C. Possibly because of this drop in temperature overnight the bats were entering the roost relatively early with all bats (about 1500) in the roost 30 mins before sunrise. The small area of hotter roof is the installed bat access slate, required under licence but not utilised as we kept the existing entrance.

New Service - Potential Roost Features in Trees

Tracks Ecology has teamed up with local tree surgeon Dave Chandler to provide licensed tree climbing services in relation to bats. We can now offer services in relation to all aspects of tree inspection and potential roost feature surveys on a flexible basis and at competitive rates. Such services include initial potential roost feature assessment, climb and inspect, activity surveys and depending on presence of bats, controlled felling as part of licensed or non-licensed method statements. If you think you may have potential issues with bats in trees then please feel free to get in contact.

Inspection of a tear cavity with a video endoscope on a Scot's pine.

Inspection of a tear cavity with a video endoscope on a Scot's pine.

Update - Bat Roost Conflict Resolution

Good news with respect to the roost conflict mitigation in Fort Augustus (see post of 23rd March 2016). The bats are back in good numbers indicating that at the very least the mitigation work was successful in maintaining the bats at favourable conservation status. We'll need to undertake further surveys to assess reproductive success at the soprano pipistrelle maternity roost and also whether or not the conflict is resolved!

Check out a short clip of bats emerging below.

Sunbathing caddisfly larvae

On a recent water vole survey in some glorious Scottish weather I had the pleasure of watching a swarm of sunbathing caddisfly larvae over lunch. It reminded me of freshwater biology fieldtrips during my undergraduate years. Check out the timelapse video.

If at first you don't succeed...

... try, try again. So that's what this pine marten did! One of my favourite parts of my job is using camera traps. Over the years I've camera trapped in various places including the Belizean jungle, the bottom of the Grand Canyon and all over the UK. No matter where I deploy cameras and for what reason, I still get excited when looking at the results. Here is some nice footage of a persistent pine marten. For my clients, camera trapping offers a very efficient and effective survey method to compliment more traditional field sign surveys. 

Bat Roost Conflict Resolution

For some time Tracks Ecology has been providing pro bono advice in relation to conflict resolution at a site in Fort Augustus where a large soprano pipistrelle roost (>1000 bats) has been causing problems in a guest house. The issues resulted in modifications to the roost being required and this week we saw the roost modifications completed (undertaken under SNH Licence).

Works underway with fine weather!

Careful consideration of the bats and how they are using the roost was required to maximise the chances of maintaining the roost at favourable conservation status. We constructed a large bespoke internal box at the location of the existing roost with entrance locations restricted but maintained in the same location. Additional measures include a summer/winter external bat box (also acted as a receptor for a small number of bats found during works), a secondary roost entrance through a bat slate and an increased number of crevice roost spaces within the main roost. 

We'll return later in the year to see if the measures have worked and the bats return in good numbers.

Cameratrap Surveys

Tracks Ecology are presently undertaking protected mammal surveys on a Highland Estate and have been utilising camera traps as an effective compliment to field sign surveys. After a two week period numerous photos of pine marten, badger and otter have been obtained along with some non-target species such as red squirrel, fox and buzzard. Camera traps have been rebaited, batteries changed and memory cards downloaded. We're looking forward to see what the next few weeks bring.

2015 Survey Season

Although we are still in mid winter with snow falling, now is the time to start thinking about your programme of works for the summer season. For the vast majority of surveys seasonal constraints are an important factor and missing survey windows could result in significant delays to projects. Tracks Ecology is already taking bookings for bat surveys. If you think that your project may require ecological surveys then please get in contact to discuss your potential requirements.