/sci/ - Science & Math

>>6707022
>>6707003
I;m aware that a lot of animal motor skills are more complex and harder to replicate then people realize, but i'd think that it could be programmed into a bot through trial and error with proper testing in a lab; an algorithm could be developed to direct the bot to find an error in the car's parts (for which the firmware would have to be constantly updated through patches for new automobiles and parts), and then it could perform a precise operation to remove the part and replace it with a new one that is compatible, probably faster and with less haggling then a biological mechanic. motor movement is not something that requires creative thinking once the formula for such movement is learned, and could be easily replicated over and over once the testing data is broken down into mathematical algorithms coded into the machine.

they already have machines that can simulate human bipedal walking quite well, and the military has a unit for carrying gear that simulates the walking of a dog