Its development and use


This blog describes the construction and use of the "Hurley Dolly". A camera dolly is a mechanised system to allow a camera to move. This is easy enough to arrange in a studio, but we wanted to use it in remote and difficult places. The dolly and track is constructed of rugged and easily-mended materials. It is designed to be simple to use when the weather is bad, the operator is tired, hungry and cold. It works. Frank Hurley was a pioneering photographer of the early 1900's. He was the photographer on Douglas Mawson's Expedition to Commonwealth Bay in Antarctica. The Dolly was first designed for use in Antarctica on an expedition to Mawsons Huts, so we named it after Frank Hurley. The Hurley Dolly. To read this in chronological order, start at the end and work upwards.


The Final Result

This is what the trip was all about. A fulldome 20 minute movie on a 4m dome at the Aboriginal Cultural Heritage Centre in Carnarvon Western Australia.

An amazing presentation, developed by Dr Peter Morse partly on the iVEC supercomputer at University of Western Australia, shows the Jangurna story of the people in the stars, plus timelapse photography of some significant sites in the area.

Photographs courtesy of Paul Bourke, Research Associate Professor and Director: iVEC University of Western Australia  and  see also

and see also the story of the trip at

The Hurley Dolly on location

Pete has been engaged to do a fulldome movie of the night sky for the Aboriginal Cultural Heritage Centre in Carnarvon.  This entailed taking two EOS 5DmkII cameras plus fisheye lenses, and computers plus two dollys – the dynamic perception MX2 dolly and the Hurley Dolly in a 4WD and driving from Hobart to Exmouth.  5,000km across Australia and half way up it.

The story is told on at which you’ll find pretty interesting.  Hard work, very hot but rewarding.

The Hurley Dolly proved to be quick and foolproof.  Also very rugged.  It just did what it was supposed to every time.  The batteries were in the box and seemed to last forever.  The switches and knobs are excellent – when it’s hot,and you’re tired (and we were!) then you don’t have to think because everything does what it says.

I’m making a few modifications now – a buzzer to tell you it’s stopped its run, a wider vernier scale for mm moved, and an ability to take over the timing using an external hand-held camera timer.  I’m also remaking the frame, using a professional bending service to make a better-looking system.

Here’s a few pictures of the Hurley Dolly on our trip. Click on a picture to enlarge it and get a slideshow.

Fully portable timelapse dolly


This one now has a portable frame.

The frame is made from a lightweight ladder, cut into 3 pieces which can be carried easily through the bush and reassembled.  The supports are 3 tripods – which are themselves lightweight and portable, and at least one will be part of the kit anyway.

If the platform sways in a strong wind, which we found was a problem with previous trials, it can be guyed using tent pegs into the ground or attached to firm objects.



HDR timelapse

Taken with Pete’s Canon EOS 5 Mk II, processed using Photomatix.

New time and motion movies

Here are some shorts done with this setup (none have any glitches, but they might show the occasional jerk with internet download or a slower graphics card).

The latest Hurley Dolly is great – it’s easy, rugged, and reliable, and folds away into a backpack for getting where we want to go.






Compare this with the previous video.  Using GBDeflicker, the video has been made smoother.  Good program.


So after all this effort I finally built the present one which seems to just work: Hurley Dolly 9.1.

By this time I have done a lot of design and building, tested and scrapped a lot of ideas, and done a lot of web research.  I knew about the guys at Dynamic Perception, with their OpenMoco based dolly system ( and (

Pete bought a Dynamic Perception setup, and I bought their Arduino shield to build one myself.  It works really well, and is very clever.  The quality of their board and the detail in their engineering is superb. If you want really good stuff, go to them.  Best I’ve seen without costing very large dollars.

However, a couple of things didn’t quite suit us with the Dynamic Perception gear. The track is only 6ft, and is made of specialised aluminium section which you can’t get in Australia (or if you can it’s probably very expensive). This track isn’t long enough for distant shots, though they are selling a 6ft extension now.   However, I’m looking for something which is adaptable so we can use onsite meaterial if necessary (like a ladder).  And, oddly enough, it’s too clever.  The problem is the wide range of settings in their menu-driven interface, and the large number of things it will do. As I said before, people get tired, hungry and cold and make mistakes.  The simpler and more rugged this gear is the better – what I’m after is knobs and switches.

But if we are doing dolly-timelapse in controlled environments, Dynamic Perception’s gear is the way to go – and I reckon we will use it too.  But we want to sling this stuff in a backpack and take off into the bush with the camera – and still get superb footage.  So I continued the original design route concentrating on a simpler approach.

What I really liked about the Dynamic Perception setup was their drive mechanism.  It’s a heavily geared 12v DC motor with a toothed pulley and guide wheels.  They have also started selling 12ft belts.  So when I was recently in USA I bought the motor and belt and made some electronics to drive the motor and trigger the camera.  Put an internal battery in the box, rigged a tensioning system for the belt, strapped on the old skateboard wheels, and here we go.  A big plus is that this will also work with nylon cord – wrap some around the pulley wheel and tension it right, and it works just as well (not so good uphill though).


Things that didn’t work

It took a year to develop the latest camera dolly.  A year of making mistakes, having ideas, building stuff and moving on.

I don’t mind this process – I spent 7 years doing scientific research and the only way to progress is to be wrong.  But you have to be wrong in the right way – you have to understand why you’re wrong and what you need to do next time.  If you don’t understand that, then you’re doing things more or less at random.  So everything I did wrong, I made sure I understood why before designing something else.  The internet helped a lot – there you can look at other peoples ideas, and, most valuably, look at what they did that failed so you don’t have to do it.  Saves heaps of time.

The way to be more successful is to make more mistakes.

Here’s some ideas that were simply daft in the first place, or failed for some other unforseen reason.


Grandma’s electric cart worked surprisingly well – but was far too fast for the dolly shots.  Nevertheless, for straight movies it looks like a good one.  However you do need a granny who is helpful, plus cart – I realise that all 3 are not necessarily available to everyone.


The golf cart not only looks silly, it doesn’t work either – too fast and too bumpy.  I then geared it down by 1/5 and put another motor on it, but it was still bumpy.  However worth bearing in mind that these golf carts have huge pulling power – they will take a pretty weighty load of golf clubs up hill, and are very well engineered.  Lots of them around – these were scrap for $5 each.


This one worked pretty well, and was robust, if a little big, but was too fast.  It’s made from a garage door opener motor and gearbox, printer paper roller, and the ubiquitous skateboard wheels (which have several grades, depending on the number of bearings in them: betcha didn’t know that).  It gave a really stable image, and I might revisit it one day.  Incidentally the idea of having lots of wheels on castors is to avoid jerks in the camera if there is any bump in the track.  It pretty well removes the effect of little ones (ie 1mm or so).


This was a clever one, pity it didn’t work, but the idea is right.  It’s a stepper motor and driver with a belt drive to 16 wheels on the tracks.  Worked well, was robust if a bit heavy, and was adaptable to any size of track (very important for what we’re doing – we want to be able to use anything that comes along if need be).

There were two problems – it wouldn’t go uphill, only on the level – the wheels slipped on the tracks.  And the belt, which was a rubber ‘O’ ring, stretched and slipped.  I could have got a better drive belt, but it was obvious it wasn’t going to work anyway so I scrapped it.  Lot of work though.


I also experimented with long track.  Ordinary black poly pipe just doesn’t work – too soft, too bendy and too hard to control.

Then I tried this green stuff. What I found out was you need support all the way, which means it has to go on the ground, or be made of something solid (like the steel section we used in Antarctica).  This is thick walled high pressure water pipe, pretty stiff but it still sagged between the sleepers.  It would work, but at the cost of a lot of infrastructure.  The ladder and beam arrangement of the West Coast trip illustrates this – we needed a 4WD to cart it around.

So what we’re going to end up with in the track department is probably a ultra-lightweight custom-made foldable track.  Either from thin steel and electrical conduit, or from a cheap light ladder.  The advantage of this is it will fold up and go on the back of a pack.  In theory.

The other advantage is that if for some reason we can’t take it with us (on a plane for instance) we can use a ladder from the location, or corrugated iron, pine studs etc.