0 – Technology – Design Methodology
Traditional actuation design starts from a set of output requirements, such as peak torque and speed, plus some restrictions on the implementation, like DC bus voltage, power limit, volume and weight. Some other implementation requirements can be added, such as comms protocol or cable management.
From there, the actuator designer usually starts making decisions. The most obvious ones being which motor and transmission to use. Those choices, despite being seen as simple and typical, might severely compromise the performance of your actuator, preventing the designer from realizing the actuator’s full potential. Is there any other way to do it? Quick answer: yes.
The long answer is a more complex, and it requires explanation about our design workflow. We decided to build the ultimate digital twin: a physics-informed, virtual representation of our motors. It includes mechanics, electromagnetics, thermals, control, electronics (control, power, sensors),… you name it.
Less than 2% error. This is exactly the gap of our model compared to our physical motor. And it´s great. It means we can use our digital twin for motor design; for control programming; for research… This is what is allowing us to create more torque dense motors, with better thermal behaviour and great bandwidth.
We didn´t stop there. Once you have the motor, why not expanding our digital motor into a full digital actuator? Exactly. We did.
We added transmission, sensors, electronics, higher level control, power and control electronics… All the necessary to achieve a full predictive model of a QDD actuator, where we can modify parameters at will on every domain and see the aftermath in output performance.
The potential of our holistic approach, combined with the capacity of a predictive virtual model, can be harnessed to provide a powerful generative design tool to explore thousands of parameter permutations and reach our goal: produce the absolute best actuators to enable physical human-robot interaction.