Industrial Robotic Arm manufacturers such as KUKA and Bosh all offer similar products, granted each one with its own positives and negatives – Whilst I was researching these robotic arms (after watching
this insane video!) I was intrigued by the inner workings of these machines, and the difficulties that the designers face when creating the design. I wanted to see how difficult it would be to design one from scratch, so I started work on the design in November 2018, this would quickly become one of the most complex projects I’d worked on at the time. I created a design brief, with the specifications of the project, and the desired final outcome.
Easy to assemble (easy tool paths) Easy to manufacturer (multiple of the same components used) Rigid frame and base Bolted to the ground Quick attachments to end of arm Able to move to any point in the surrounding area without hitting itself.
As with any project I started by doing some research – I quickly realised that this project was going to need industrial bearings, so I looked into the difference between the different sizes, max load, mounting type, etc. And eventually decided on the RU85-UU Cross Roller Bearing – as it offered the best combination of all of these for my use case.
With the new knowledge of the bearings and my extensive previous uses of motors I started with a very basic drawing on Paint, the idea for this drawing would highlight how each component would move, it also allowed me to refer back to it when I was creating the individual components. From Here I started work on modelling some basic components that I’d later use:
NEMA 17 Stepper Driver
22mm Skate Bearing
Starting to Design
From the start I knew that this arm needed to be bolted to the ground, when I first started the project I intended the arm to be able to ‘snap in’ to the base plate however I quickly realised that this was very unreliable so I settled on an 8 Bolt system. This is what I started with:
A Section analysis of v1, showing the different bolt-in points for the bearing.
After a couple of attempts, I was left with this baseplate – it allows the builder to quickly assemble the base – and has easy cable management for the builder to use.
I then started work on the main bulk lifter of the arm, I knew that this component (asides from the base plate) would be under the greatest load, and so I had to design it to be able to withstand a lot of weight, I also had to figure out how to create a stable lifting point, I realised that if I used a lever like the design to increase the amount of weight. I also started playing with using a Pneumatic Actuator.
In my design I used a Pneumatic Actuator connected to a pivot point, this will allow me to push and pull great forces without having to use extremely heavy-duty systems, which would reduce the overall cost of development and manufacture. From this point, the design started to become more and more complex, partly because I needed to start to think about cable routing and Toolpaths, as these were key features in my requirements for design. To do clean cable routing I had to figure out cable and piping diameters, to create correctly size wholes.
I should note that it was around this point that I started making every ‘Joint’ into an active joint – meaning that I can grab the end of the arm and the computer knows how far and how to move the arm into the appropriate position without breaking the model.
Another element of this design is that it would be easy to manufacture. For this reason, I created this segment that could be used twice per assembly, this means that I was able to cut down on the number of specific parts made for this project. This meant that the entire assembly could be built with only 6 custom parts, the rest of the components are off the shelf, which means that it is easy to replace and maintain.
As mentioned above I wanted to make sure that I can use certain components multiple times to reduce the complexity of the assembly process. So I created this part, they are standardised and allowed me to use them repeatedly to create my final model, effectively acting as a building block.
One of the design brief elements was that I needed to ensure that the tool paths were easy to get access to – this purple to green segment-best demonstrates how I was able to accomplish this. The RU85-UU Bearing has 8 embedded bolt points which meant that I was able to easily create a flush point that I just needed to make accessible to the final assembly.
I used two of the Green Elements, and a single purple piece to create the main body of the arm, each joint has 360* of movement allowing for 4 axis’ of movement just from those three parts, and once I attached this assembly to the bottom assembly that I’d previously designed, it gave me a further 2 axis’ of motion, in total giving me 6 axis’ of movement, allowing full control.
After I was happy with the range of movement I created a final end piece, this would easily allow a user to attach custom tool heads easily.
Finally, Once all assembled I was left with this model! I’m extremely happy with my final design, It was an extremely complicated project and in a field that I hadn’t had a huge amount of experience in before, I defiantly learnt a lot during the development process, and I think that this shows throughout. I took on this project as a challenge, to push myself and my design ability and to start considering things that I’d otherwise overlook, and for this purpose, I think I excelled in this. Looking back, there are definitely things that such as designing the parts to be easier to CNC mill, I’d also look into stronger Pneumatic Actuator.