rationale behind max_delta_angle_factor

[anahrendra]

Hi! I have a question regarding the temperature-based delta angle clipping part of the depolyment code

CLONE/deploy/g1_server.py

Lines 667 to 674 in edffec4

	self.joint_temperature[:] = np.array([msg.motor_state[i].temperature[1] for i in range(HW_DOF)])
	# self.env.max_delta_angle_factor[(self.joint_temperature > 60) & (self.joint_temperature <= 80)] = 1.0
	# self.env.max_delta_angle_factor[(self.joint_temperature > 80) & (self.joint_temperature <= 100)] = 0.5
	# self.env.max_delta_angle_factor[(self.joint_temperature > 100) & (self.joint_temperature <= 115)] = 0.1
	# self.env.max_delta_angle_factor[(self.joint_temperature > 115)] = 0.0
	# self.env.max_delta_angle_factor[(self.joint_temperature <= 60)] = 1.0
	factor = 1 - ((self.joint_temperature - 80) / 35).clip(0., 1.)
	self.env.max_delta_angle_factor = factor.astype(np.float64)

My guess is that it is meant to handle nonlinearities when the motor gets hotter. But do you have any reference for the rationale behind this choice of implementation? Thank you!