Fully automated end-of-line test bench for BMW eDrive

The latest generation of electrified drives is produced at the Dingolfing site in Germany, the BMW Group competence center for e-drive production. Here, KUKA implemented a fully automated test system for testing up to eight different drive variants. The KUKA team of experts, based at the Bremen site, invested around two years in the design, construction and ultimately the successful commissioning of the eDrive test system.
More information about assembly and test systems for the powertrain: www.kuka.com/en-us/industries...
www.kuka.com/BMW_eDrive_test_bench

[@ajpt4336]

The purpose of both functional and performance tests is to calculate the efficiency of the e-drive system by comparing all results of measuring output mechanical power of the system such as torque, speed, acceleration, and energy with a variable input electrical power, then the variable and sudden external mechanical loads are applied to the system to measure stability and response of the system to all external changes and shocks in conjunction with the measurement of vibrations, noise and heat resulting from all parts of the system during tests mediated by NVH sensor and then all test results are displayed in the form of several 3D differentiable curves to calculate the points where the efficiency of the system reaches the highest value to note if the system satisfies the required calibrations so that engineers can modify the design of the system by a special simulation program to reach to the optimal design.

[@jashanth.ngowda4149]

Amazing

[@hartalex69]

Excellent!

[@theeaxleguy3313]

It's very interesting to see that the two different BMW eAxles (one mounted flat, with the inverter on top, and one mounted vertically, with the inverter on the side) are tested on the same automated equipment. This flexibility is a nice feature. I am assuming that one of these eAxles is for the front of the BMW iX (mounted vertically under the hood), while the other eAxle is for the rear (mounted flat, to reduce the impact on the trunk space). These might both be HEAT-L (i.e. HEAT Large) variants, or perhaps the rear one is a HEAT-L and the lower-powered front eAxle is a HEAT-M (i.e. HEAT Medium).

[@tk7836]

Interesting video. Do you do the parking lock and drag torque test on the same motor & sensor configuration like the functional test? Was not completely clear to me in the video. In that case, would that not be difficult for the drag torque measurement sensor accuracy? (2000 Nm each axle for functional test…). And also long cycle time? Just wondering. Very impressive test rig, I really like the flexibility with the two robots, for different types. Also the NVH sensor application with the robots looks good, allows a lot of flexibility for different types in testing.

[@tk7836]

Sorry the editing function seems not to work. I meant 2,000 Nm per output shaft, if I understood the description correctly.

[@WeisM3]

I don’t believe that they are the same. I can see that they are using our torque limiters in this assembly. Our torque limiters are preset so the two separate tests will have two different torque profiles. One test profile may trip the torque limiter while the other doesn’t. This is just my theory.

[@williamhuang5329]

Hanzhen harmonic drive gear , industrial robot arm gear box

[@eletronika2]

Esplêndido

[@aegystierone8505]

the music is really unnecessary, people want to hear the robot noise.

[@KUKARobotGroup]

Thanks for the feedback!