No much of a review, more of an unboxing. A honest review is always good for the person who is spending the cash. I was expecting you to measure it's payload, repeated accuracy- is it really 0.2mm repeatability? Does it have any backlash on any of the axis when the arm stops moving? How fluid is it in comparison to industrial robot? How much better is it than a servo arm (likes of LewanSoul Hiwonder a servo DIY arm with similar programming)? What is the quality, finish like (is it cheap plastic construction?). Your verdict on long term reliability and use. Any other disadvantages / advantages. One of the pain points I see straight away is in teaching mode- you have to move the arm via pendant or computer software to exact points and then record this in the software. This is a extremely time consuming process as it can be hit and miss to get the exact coordinates. It would have been better if you can just grab the arm and move it to the specific point and record. This can not be done with the Mirobot as there are gears connected to the Vic-Tech stepper motors used in the arm. Maybe in Mirobot version 2 they can change this as well as update any other hardware issues. Plus make the pendant display bigger (OLED) with removable batteries. The professional kit sells in the UK for £1870, that is a lot of money for a arm which picks up and moves things around on your desk. I think the novelty will wear out fast. I can see a use in education to teach programming. I think this should be priced at £600 max, even at this price it's a lot for a plastic arm. It looks like Wlkata are ripping off educational establishments for what it is.
I'll be honest, i love the arm, and i understand why it costs 2k. But it still hurts me that even if the inverse kinematics is done and everything is packaged nicely into a single embedded controller, that they chose that controller to be an ATMega2560, for the love of god put it on STM32. Which also leads into the next limitation, since it doesn't seem like there is any specifications regarding encoder resolution. So we have a nema based feed forward robotic arm built on a MCU that only remains relevant thanks to it being the lowest cost step into embedded programming. While the rest of the arm looks very classy, seeing it shake when bumping the table and having no real evidence to believe it actually knows where it is (feedback). We also know it can't perform precise tasks both due to mechanical compliance and the lack of any more advanced control schemes. FOC/Feedback/Gravity compensation etc, The motors should be at least Nema converted to a servo. Needless to say this is on that fine line between toy and functional product that is made for educational purposes, but not for students. But the price is right, there is fuck all market for a product like this so of course the cost will be high, they aren't selling like pancakes.
