Delta X S Prototype

Delta X S D800 Working space

The major improvements in building the Delta X S robot prototype include transitioning from Spherical rod end ball joints to Ball-and-socket joints and replacing planetary gearboxes with pulleys and belts for gear reduction

We have extensively searched for suppliers for the delta robot ball joints, but none of them met our criteria of being both lower cost and having the lowest possible sagging. We successfully fabricated ball-and-socket joints for the Delta X 2 using SLA 3D printing technology, but it was not feasible for the industrial-grade Delta X S robot. The 3D printed joints would inevitably break and not meet the required food safety standards. We considered options such as 3D metal printing or milling solid stainless steel, but the costs were significantly high, even when purchasing joints from companies like IGUS. Therefore, we came up with the idea of using solid BOM plastic and milling it. The results were truly remarkable. It took us over half a year to find a solution to reduce sagging caused by the ball joints.

Spherical rod end ball joints and Ball-and-socket joints

And the most time-consuming aspect for us was the gearbox.

We went through nearly 10 different motor and planetary gearbox combinations, but none of them met the motion requirements of the delta robot. The gears would continuously break and wear out within minutes of operation.

It would cost around $5000 for a reliable set of motors, gearboxes, and drivers for the robot. This was truly insane considering our goal of selling the Delta X S within the $4000 or lower price range.

 We had to come up with our own new design to reduce the manufacturing cost for the Delta X S. That's when we opted for the pulley and belt-driven gearbox.
 
Little did we know that it would take us an additional 3 years to address the issues related to the gearbox before the Delta X S robot could operate smoothly in an industrial environment.

Well, I must say we tried dozens of different belts available in the market and made numerous design changes along the way.

Planetary gearbox and pulley gearbox

First test run of Delta X S V2

With this change, the moving base of the robot has reduced its error. However, a new issue arises as the belt tends to loosen after a short period of continuous operation.

In this version, the robot has achieved 70% of our expectations. The remaining 30% has posed significant challenges for us in the coming years

After two months of improvements, we have developed Version 1.0.

It is now capable of functioning in low-intensity environments outside the lab, so we decided to test it by selling it to a customer in the United States. The customer is a group of Makers who have a passion for mushroom cultivation. They have successfully raised several million dollars in funding to build an automated farm using robots.

Delta X S V1 3-axis with a large Z-axis travel

We have sent them 3 Delta X S V1 3-axis robots along with a special gripper for holding plant pots. They are quite busy with agricultural engineering, so they haven't responded to me regarding the use of the robots. I lost contact with them after the COVID pandemic. Hopefully, they are doing well and find the Delta X S robots useful.

Gripper for holding plant pot

Below is a Delta X S V1 5-axis robot. It has a large control box that contains drivers, power supply, and motion controller.

This version still has many limitations and drawbacks. It has excessive backlash due to inaccuracies from the gearbox and ball joints.

Despite changing many suppliers, we have not found a low-cost yet high-quality gearbox supplier. The gears in the gearbox immediately wear out after a few minutes of continuous reversing. These types of gearboxes may not be suitable for the fast motion of delta robots.

My initial goal was to create an industrial delta robot model with an official selling price of only $2000. However, after considering the available materials and equipment on the market, this turned out to be unattainable because they couldn't meet the durability requirements of an industrial product. Therefore, we had to change our goal for the upcoming versions.

Here is the desired configuration for this V1 version. However, the expected achievement is only around 60%

Delta X S V0.9, our industrial delta robot, was initiated in April 2020.

Building an industrial delta robot is our long-term goal, and every small step we take is geared towards achieving that.

We initially released Delta X 1 to raise funds for Delta X S research. However, the business of Delta X 1 failed, and we faced the risk of bankruptcy.

Surprisingly, an American guy offered us financial support and requested us to research an automated ice cream vending system using the Delta X robot. Since Delta X 1 was not capable of fulfilling this requirement, we had to focus on the research of Delta X S with a limited budget.

To have more control over the processing of aluminum components for Delta X S, we also explored a 3-axis CNC machine capable of milling aluminum. This was our first attempt at building a CNC machine, and its performance was not optimal. It allowed us to create Delta X S parts, but with significant inaccuracies.

When the robot operated at high speeds, it exhibited considerable oscillations. We also upgraded the electrical circuit of Delta X 1 to handle 24V motors. This circuit was used for testing the program logic and lacked the typical industrial protection mechanisms.

After the initial test run, we observed significant inaccuracies in the robot's movements. The root cause was identified as backlash in the gearbox, which failed after a few minutes of running the demo program.

From that point onwards, until the next upgrade version, we replaced numerous motor and gearbox models from various manufacturers in our pursuit of improvement.

Overall, this version has met about 70% of our expectations. We invested $1500 in building the CNC machine and $1200 in creating the prototype of the Delta X S robot V0.9.

And we have only a few hundred dollars left in our account to sustain ourselves during the COVID closure period. It's quite challenging!

This is your heading text. 3