SKY CAM an Aerial Camera Solution




About: GCSE Electronic Products Student Year 11 UK

Intro: SKY CAM an Aerial Camera Solution

This Instructable will walk you through how I made a remote control (Semi Autonomous) Cable Cam system for my GCSE Electronic Products Project at School and hopefully enable you to be able to create your own system! This is intended as a rough guide on the principals as each system is different depending on the requirements. For this project you'll need a reasonable understanding of electronics and CAD CAM (Computer Aided Design/ Manufacture) although don't be put off as simplified versions can be made.

The Problem:

  • My client needs a system to get aerial film of a variety of activities and events.
  • The Problem is that where Drones/UAV’s would typically be used to get this footage, it is unsafe and impractical to use these over people, inside, or in typical sporting terrain such as wooded areas or a sports hall, because of the danger of injury should the system fail and limited space can make It impossible to operate such systems.

Based on this I set a Design Brief:

  • Design and make a product to capture aerial footage using a safe and cost effective system that can be remote controlled and move between two fixed points.

As most Commercially available Cable Camera systems come in at about the $4,000 plus mark. I wanted to make a system that would make this kind of advanced camera work available to more creators and hobbyists on a tighter budget.

What you'll need to complete this project :

Access to a 3D Printer (Housings)

Access to a Laser Cutter (Main Body of the rig and for the Control panel cutting and etching)

Be able to make PCB's as almost all of them in this project are custom designed.

Additionally these are the main specialist components that I used:


Illuminated Green PTM switches x3

Switch Covers for the above x3

4 Axis Microswitch Joystick

Membrane Switch (The ENT menu scroll Button)


Wheels x3

Dyneema Cable (Choose Length depending on where you plan to use the system)

Yellow Flight Case ( For the controller, although any enclosure could be used)

Step 1: Overview

The Cable Cam consists of three main parts:

The Actual Rig (The part that carries the cameras and drives along the cable)

The Controller (Contains a Microcontroller and an RF Transmitter)

The Cable (Supports the rig and allows it to be run between any two reasonably sturdy points)

Step 2: How It Works

As you can see in the pictures above the Rig Relies on friction in order to transfer drive from the wheel onto the cable (Green Line). It can be difficult to achieve the right balance of friction so I used the below methods to achieve optimal tension and friction.

Primarily the arrangement of the the wheels forces the cable down and over the drive wheel as seen in the diagram above. This is a very good method as it allows the two outer wheels to take the full load of the rig onto the cable (Meaning you can mount reasonably heavy cameras or equipment onto the rig) be sure to READ STEP 7 before trying to use your own system!

However the three wheel arrangement relies heavily on the cable be at a very high tension which is ideal and easy to achieve with my rigging method however it may not always be at the optimal tension. To cope with this the load bearing wheels both sit in a slot system that allows them to be moved up and down to vary the tension in the rig. It also acts as a basic safety system- If the cable becomes over tensioned for any reason then the out rigger wheels slide up to reduce pressure on the rig and drive wheel, hopefully preventing damage to the motor.

So when you're designing your own rig using the tri arrangement of wheels is an excellent method for ensuring drive onto the cable.

Step 3: Controller

To make the Controller you need to start off by getting the measurements with some digital callipers and making a rectangle in CAD (Fusion 360), I then laser cut it in card. This will allow you to get the measurements correct as well as letting you position all your components before you cut in your final material.

For My control panel i cut it in 3mm Dual Layer Laser plastic which is how I got the engraving's to be white. one issue with this 3mm plastic is that it's pretty flexible. To counteract this I used 5mm black acrylic with large cut outs in the support the back of the panel as can be seen above. Then in each of the four corners I drilled a 3mm diameter hole and put a machine screw through the two plastics and secured it with an M3 Dome Nut on the top of the panel which holds it together nicely. It also adds to the rugged aesthetic of the control panel.

One advantage of using a flight case similar to the above is that your panel can be a friction fit. Just push it in tightly and it will stay in, however if you need to remove it for any reason you can still click it out with a bit of force!

Step 4: Electronic Systems

There are two main Systems in this project which are outlined below. You can also get an understanding of how they integrate with the flowchart above.

The Rig Side:

  • RF Reciever
  • Microcontroller Board
  • Motor Driver

The Controller Side:

  • Control PCB
  • RF Transmitter

The Motor driver is a Transistor Based H Bridge that works by switching on alternate pairs of the four transistors that allows the motor to be run: Forwards, Backwards and also act as a break by feeding back EMF into the motor. Instead of using relays I used an Optocoupler (16 Pin 4 gate) to interface the PICAXE Microcontroller with the transistors ensuring that the Microcontroller isn't damaged by the higher current of the motor driver.

On the rig at each end there is a micro switch that allows for the rig to know its position along the cable at all times, allowing for autonomous features and also stops it from hitting the end of the line. It can determine it's location because during the start up procedure it runs the maximum length of the cable and records the end and start points. Then it can calculate its position at any time based on how long it been moving along the cable

Step 5: Drive System

In oder to be able to drive the rig at high tensions and at reasonable speed you'll need to use a high torque geared motor. I used an electric screwdriver motor and designed a housing for it in CAD (Autodesk Fusion 360) works really well for me. Above you can see that I included cable ducting and air vents, as well as mounting holes.

The measurements need to be extremely precise otherwise you may experience gear box failure. (I did in my initial testing phases, there was some excess space inside the housing that allowed the gearbox to come loose and fail so watch out!)

Step 6: Motor Driver

I decided that i would build a custom motor driver for my project that would enable me to interface the motor with my Microcontroller and run it forwards and backwards. Additionally in my final system I was also able to use back EMF to act as a brake on the motor. See above for the Motor Driver research and incremental development. After my research I finally used FET's interfaced with an Optocoupler as the pictures explain!

Step 7: Rig

Step 8: Software

The system has two Microcontrollers one on the rig and one in the control panel.

The Code for all the systems is written in BASIC on the PICAXE program editor.

If you wish to replicate I advise you look to the flowcharts as this will allow you to implement it on any platform regardless.


The original code shown here was an early stage development code and has been removed as it is unhelpful.

Step 9: Finishing Details

  • To give the product a professional finish I was able to use a Roland Sticker Cutter (Dr Stika) to cut Vinyl Sheet into text for branding.
  • Additionally you can use strips of tape to indicate the correct orientation for the power packs on the power unit. This lets you easily switch out the battery packs without getting them the wrong way up.
  • I Polished the Aluminium spacing tubes on a bufffing wheel to add the the sleek aesthetic of the device. this only takes a couple of minutes and gives a really nice finish


  • Try to polish the Aluminium tubing before you cut it too length as it will save your fingers from the buffing wheel!

Step 10: FILES:

Microcontroller Contest

Second Prize in the
Microcontroller Contest



    • Optics Contest

      Optics Contest
    • Plastics Contest

      Plastics Contest
    • Audio Contest 2018

      Audio Contest 2018

    28 Discussions


    3 months ago

    It is almost unimaginable projects like this that makes me feel utterly inept!


    4 months ago

    Wow, your school had a far higher budget for GCSE Electronics than mine did. We had a CNC machine and thought that was pretty top-end.


    4 months ago

    This is a really great design, I love it..!! Consider this from an experienced engineer; Most shortcomings in a design start at the beginning, usually long before the first line is drawn.

    My first thought when I saw this was stability. A trolley is a very difficult platform to stabilize. Take a quick look at passenger trolleys.

    One possible solution is two, parallel lines. A second improvement (to reduce bounce) is avoid high tensile steel lines because they are great springs with very little internal friction. I would suggest a rope that has a lot of internal friction, climbing rope for example. Climbing rope is made to minimize bouncing in the case of a fall and hence, reduce the chances of more injuries. Thirdly, consider moving the main trolley motor off the trolley and onto one of the fixed points on the end. In general, the smaller the mass you have, the easier it is to dampen. Also, less mass = less tension.

    To further reduce the trolley mass, you could put two motors at one of the fixed points and control both translation and rotation with them. Take a look how single sheet X-Y plotters are made, it's very clever. Consider what's remaining of the trolley in this case....a small, light frame with a 5 light pulleys and a small tilt mechanism, otherwise practically nothing...

    An added benefit of reducing the complexity/mass of the trolley is cross sectional area reduction....i.e. resistance to the effects of wind...

    2 replies

    Reply 4 months ago

    Some really great points here. The reasons that I avoided separate motors is that the system was required to be extremely potable so having it self contained was an advantage. I used Dyneema cable for the main line, avoiding as you said steel cable for that exact reasoning. And finally use of two line would have increased stability, however due to the increased setup time e.t.c i would personally opt for an electronic gimbal. However these are all really good points that are worth considering in any future designs!


    Reply 4 months ago

    One other stabilization method, gyro's....(is that what you mean by "electronic gimbal"?)....cost unknown but you may extract some really good ones from expired aircraft instr.

    Also, if you can find three (you may only need one or two) small, high quality, balanced motors w/ball can safely run them 50-100% overspeed...possibly from old tape recorders (disable the governors)...

    Again, love the design....excellent job....

    I would be interested to know how it performs on a gusty day with a long line...


    4 months ago

    Hey cool stuff!! Looks really good, but why dont you show any footage from the actual camera riding it?? Thats the interesting part...

    1 reply

    Reply 4 months ago

    Only because of the shabby camera quality in the action cams. The footage is pretty stable at lower speeds as lots of the vibration is dampend by the cable. If i manage to get a better camera that I'm allowed to mount ill be sure to upload some footage!


    4 months ago

    This is a really great project and I enjoyed reading about it; when I first saw it (in an email instructables sent me), I almost passed it up, thinking "well, this isn't something I'd use" - but something told me to open it up and take a look, and I'm glad I did.

    You've created an extremely well thought out and competitive system comparable to professional rigs, and at the same time limited expense; while in your design you note that the cost seemed high for a prototype, and that manufacturing at scale could lower the costs, even at your prototype's cost (minus your time and expense), it's still far cheaper than retail costs of even the "hobbyist" system you mention.

    I note that you chose a couple of options that would likely make for better manufacturing, but if someone were going to replicate this on their own, I would argue for alternatives:

    1. Choice of microcontroller - I would personally have gone with an Arduino for prototyping, then switch to a Nano, a Pro, or just a bare ATMega328 for the final product. All are very easy to work with, and can leverage a vast level of available source code and community help to achieve the implementation.

    2. In addition to the above, I would have selected to go with a pre-built motor controller, likely something like an RC ESC; the Arduino has an available Servo library that makes interfacing to RC devices simple, plus you'd likely want to add a controllable camera gimbal mount (I didn't notice one?) - and such a mount could be easily done with RC servos as well. Using RC devices also eliminates a bunch of custom electronics, which would likely make for a more robust system, as well as being easily replaceable in the field.

    3. Camera mounting, as noted, could be done with an RC servo-based rig, but another option would be a brushless-motor stabilized gimbal as is available for drones; these tend to be more complex to interface with, but since most drone controllers are open-source (and more than a few Arduino compatible), that code could be lifted from such projects for integration into a cable camera system.

    Again, the above would be best for "one off" or prototype implementations of this system, but parts could be used in a manufactured system as well; if one were going for lowering the costs, then a PIC microcontroller would probably be the better solution - but RC parts should always be considered even for a "commercial" solution, because of their inherent ease of interfacing, robustness and their wide availability out in the field (just hop down to your local hobby shop).

    I wasn't able to locate the cable you used on Amazon here in the US - but it seemed very similar to reflective paracord, which comes in a variety of lengths and colors. Such cord could easily hold the weight of this system.

    Thank you for sharing your experience and project; even though I may never have a use for such a system, I enjoyed reading about it and seeing the effort and thinking that went behind it, and I found some interesting takeaways I could incorporate into my own future projects.

    1 reply

    Reply 4 months ago

    Hi Andrew,

    Thanks for your feedback, unfortunately most of this I would have liked to use however as I'm a student (16) I had to complete the project within the somewhat restrictive nature of my course which is why some of the solutions/design is fairly convoluted. I fully agree with your alternatives and would join you in strongly advising someone who is making this without the scrutiny and regulation of an exam board to go for these easier options!


    Question 4 months ago on Introduction

    Excellent design and documentation. Questions:

    Was the gearbox from the electric screwdriver?

    How did you connect the gearbox output shaft to the drive wheel? (pulleys shown all have an internal bearing)

    Looks like you have to remove the support wheels/ shafts to attach the unit to the cable. I would think that a quick release pin here would be helpful.

    Does it have enough friction on the drive wheel to operate on sloped line? Would think that you would need more contact angle on the drive wheel or even a multi-reeved system.

    1 more answer

    Answer 4 months ago

    1) Yes thats it a 3.6v electric screwdriver with a planetary gear system just off amazon!

    2) I had to both drill out the hole by 0.5mm for the drive shaft and that allowed for a secure friction lock and then I simply used superglue to lock up the bearing and added epoxy just to be safe as its a high load part.

    3) because of the nature of the system and the way the rig is set up you can just feed the cable through, but you can actually simply release the load bearing wheels then slide the cable past the drive wheel and on to get it on.

    4) Yes, there is absolutely no problem in terms of friction which i was pretty suprised with, the three wheel system really works very well and the sliding outer wheels allow you to adjust accordingly. But theres definitely enough friction, from time to time its been too much with a highly tensioned centre cable so I've had to use the adjusters to let some slack into the system!


    4 months ago

    Neat project. you made a lot of interesting choices in the design, and it looks pretty classy. There seems to be a problem with the schematic. The two opto diodes on the right side of the motor drive bridge appear to be backwards. Reverse those two, and three wire operation should give you left, right, brake, and off with no short circuit conditions. I like that.

    1 reply

    Reply 4 months ago

    Yep, Thanks for the heads up thats just a schematic error I'm afraid made in the early development but thank you!


    4 months ago

    Note to the author and others who are thinking about replicating this project:

    The code behind this project is suspect; in the course of copying and re-formatting the code, I found several inconsistencies and probable errors which leads me to recommend an audit and refactor before usage in anything more than a prototype.

    Issues found consisted of:

    1. Numerous and copious use of "goto" statements - this is bad no matter how you look at it, as it leads to (and this code is a good example of) "spaghetti code".

    2. Numerous examples of nested gotos and gosubs intermixed; this can lead to stack overflow issues if or when a return is not completed for it's complementary gosub, which could cause the system to go into an unknown state during operation.

    3. Little to no rhyme or reason behind sub/routine label names or variable naming.

    4. No comments or other code documentation, which will make any refactoring or maintenance of code much more difficult.

    Many, most - possible all - of the above issues may be caused by a poorly implemented code generation system. I only say this because the code looks generated in some fashion, due to the weird naming conventions for the labels, and for the structure of the code itself. While I don't believe the author is a software engineer, even someone with only the basics of knowledge of coding wouldn't do some of the things I saw in the code (such as double labels for routines, when a single label would suffice, or a labeled routine that would goto another labeled routine immediately below it, or gotos into a routine that has a final return statement, but seemingly without a corresponding gosub in the execution chain - though given the convoluted nature of this code, I could have missed something here).

    Also, note that there seems to be three separate pieces of code - in order:

    1. Code that controls the platform (the camera rig on the cable).

    2. Code for the remote controller box/transmitter.

    3. Some short piece of "testing" code (not sure what that's for).

    1 reply

    Reply 4 months ago

    Thanks for the reminder Andrew. The code that was up was actually very early stage rough development code only intended for use on bench prototyping :) I've removed it as its clear its only causing confusion. If people are going to attempt to recreate the system i'd advise they address the flowchart diagrams for a rough understanding of how the system is divided! But Thanks for the clarification!


    4 months ago

    I don’t know why you went to all that trouble making that crazy rig just to use a GoPro!

    3 replies

    Reply 4 months ago

    It's not just for GoPro's you can put any camera up to a few kilos onboard!


    Reply 4 months ago

    What does it use for the actual mount? Is it a 1/4-20 bolt like a tripod? Or is it some sort of clamp? Because if you use a 1/4-20 it should be able to hold any camera the bearings and cable can support.


    4 months ago

    Can you post a finished time-lapse or some b-roll shot with it?