Assuming you have a list of numpy arrays, you can concatenate them and take the sum on batched axis and finally clip the resulting numpy array in order the values summed to be maximum of 1.

resulting_arr = np.clip(np.sum(np.concatenate(list_of_arr, axis = 0), axis = 0), a_min = 0, a_max = 1)

# list_of_arr = [np.array([0, 1, 0]), np.array([0, 0, 0]), np.array([0, 1, 1])]]

# resulting_arr = np.array([0, 1, 1])