Patients with respiratory weaknesses due to neurological disorders or aging often face usability challenges with conventional mechanical ventilators, which directly move air in and out of the lungs via the ventilation mask. To address these challenges, Exo-Abs was developed to support a wide range of respiratory functions through synchronous abdominal compression. Its control previously relied on continuous full-phase measurements from multiple sensors to ensure control performance across a wide dynamic range. However, accommodating long-term usage scenarios remained challenging, as practical features such as allowing breaks between breaths or adapting to different installation environments were limited. Here, we present a user-centered solution designed to address these real-world usage conditions. Although consecutive compression-and-recoil cycles are commonly considered essential for this type of assistance, we found that well-synchronized abdominal compression with Exo-Abs can immediately augment the corresponding breath, when applied above a certain intensity. Based on this finding, we proposed an exhalation-synchronous control strategy for the system that involves strict control policy over the exhalation phase (compression) and sparse control policy over the inhalation phase (release). A streamlined sensor configuration was also implemented to improve use scenarios, allowing users to take breaks freely and supporting long-term use. To evaluate the improved practicality of Exo-Abs, we conducted an experimnt in which the device was used in place of a conventional mechanical ventilator during prescribed respiratory therapy sessions for hospitalized patients. Notably, all participants were able to use the system for up to approximately two hours, demonstrating the feasibility of the proposed control scheme for long-term usage. The efficacy of assistance was evaluated by utilizing the mathematical model individualized to each participant. Results for primary respiratory performances showed an average 23.25% increase in the peak volumetric flow rate per breath (ranging from 13.99 to 57.81% depending on the user) and an average 19.46% increase in the maximal volume of air moved in and out per breath (ranging from 7.23 to 45.60% depending on the user). During assistance, Exo-Abs applied between 76 and 91 N of compressive force synchronously to each breath. Secondary analysis based on individualized mathematical models showed an average increase of 1.80 cmH$$_2$$O in mean pleural pressure per breath (23.44% of their spontaneous pleural pressure; ranging from 7.99 to 43.93% depending on the user) and an average 0.07 J increase of the mechanical work per breath (23.49% of their spontaneous work; ranging from 8.22 to 45.35% depending on the user). This study demonstrates that Exo-Abs can enhance respiratory performance in patients with weakened respiratory muscles, even in long-term usage scenarios. Along with the simplifications in the control policy and user interface, Exo-Abs has been shown to provide effective respiratory assistance over various breathing patterns and respiratory training contexts. The contribution of this study extends beyond demonstrating a user-centered system with improved usability and practicality, as it also establishes an objective evaluation framework for respiratory biomechanics.