When the injury affects the brain areas involved in the segmental control of the trunk, children with CP could present difficulties in achieving unsupported sitting due to control deficits of low thoracic and lumbar regions that disrupt posture and balance [15]. In this context, a few patients could benefit from external support, such as robotic devices, targeting impaired postural control. Santamaria et al. [16] delivered a postural task-oriented training through a motorized cable-driven belt, called Trunk-Support-Trainer (TruST), placed on the CP child’s most-impaired trunk region. The TruST provides maximum trunk displacement in the reaching task to promote postural reactions. In fact, it can force the patient to develop active trunk movements. This training produced clinical improvements in GMFs, as well as in postural and reaching control, allowing children to sit independently. Abidin et al. [17] studied the effects of a combined approach using RAGT and standard physiotherapy. The authors found that RAGT, as an adjunct treatment to physiotherapy, was useful in promoting trunk control, sitting balance and posture in non-ambulatory CP. RAGT using the Lokomat device has proven effective in enhancing cortical plasticity and cerebellar-motor connectivity through augmented sensory and proprioceptive feedback [18]. In line with this hypothesis, it has been confirmed [19] that training with the electromedical device Alter-G can favor plastic changes in the brainstem, cerebellar white matter and vestibulospinal tract, producing permanent postural and balance improvement in CP children [20]. The mechanical horse-riding simulator (HRS), an emerging type of intervention based on hippotherapy and consisting of a robotic device with a dynamic saddle that imitates the movement of a horse [21], is of particular interest in this field. According to a systematic review [22], HRS seems to be effective in improving balance in subjects affected by CP, especially in terms of anteroposterior, medial–lateral weight shifting, trunk decompensation, inclination and pelvic torsion. In fact, the riding movement produces soft, rhythmic, and repetitive patterns, simulating the pelvis movements during normal human walking [23,24,25]. These kinds of repetitive riding movements could improve postural coordination through the stimulation of balance reactions thanks to the effort to maintain the center of gravity inside the support base [25,26]. In a single case study carried out on a child with CP [27], HRS was also effective in increasing postural muscle tropism, including internal oblique, external oblique and lumbar multifidus, with static and dynamic stability. Indeed, the use of HRS could be a valid device to manage hip adductor spasticity, which worsens motor functions and postural control, as confirmed by Hemachithra et al. [28]. The authors showed that HRS was effective in decreasing the adductor spasticity and improving the abduction range of motion in the hip, suggesting its use within physiotherapy intervention. Recently, Jung et al. [29] investigated the use of HRS combined with virtual reality (VR) in a sample of preschool- and school-aged children with spastic CP. The authors found that the combined approach was more effective in improving GMFs, balance control, and mobility without serious adverse events, rather than only HRS training.

In patients with CP, gait is often accompanied by reduced pendular synkinesis of the upper limbs, stiffness and imbalance, which require more energy and effort rather than in healthy children [30]. The development of RAGT attempts to achieve a correct motor function since robotic tools can provide high-intensity, repetitive, task-specific, and interactive training. Generally, two types of robotic gait devices can be distinguished in clinical practice: end-effectors and exoskeletons. End-effectors are stationary devices that reproduce gait trajectories through footplates guiding feet (i.e., GE-O System) and the swinging phase of gait. Exoskeletons are wearable devices for both over ground (i.e., Ekso) and on a treadmill (i.e., Lokomat) walking. The main difference between these two robotic gait devices is that end-effectors only act on the distal part of the body (i.e., feet), while exoskeletons act on the main joints of the lower limb [31].

With recent advances in technology, RAGT has become more readily accessible and available in rehabilitation centers. Evidence suggests trends for improvement of muscle activation after active training performed with high intensity and rate of guided movements [32]. According to Wallard et al. [33], it seems that RAGT could promote new dynamic strategies of gait, which are characterized by more appropriate control of the upper body in addition to improvement of the lower limb kinematics. However, it has been pointed out [34] that RAGT is no more effective than standard physiotherapy, especially in the case of passive gait training without the active engagement of the patient necessary to promote motor learning. This is why active walking training (above all when it is used robotics plus VR) could enhance cortical activity and plasticity due to higher attention required with a better modulation of gait speed and steps, as confirmed by Yazıcı et al. [35]. The degree of gait impairment is another factor that cannot be underestimated. In fact, it has been found [36] that RAGT, using the Robogait (Bama Technology, Ankara, Turkey), promoted improvements in the standing and walking abilities only in mild to moderate CP children. Notably, wearable robots facilitate an appropriate alignment of the body axis during weight-shifting movement, improving both static and dynamic balance control and, therefore, gait function. Indeed, it seems that exoskeletons could be more effective for dynamic balance than tethered-type robots [37]. A recent systematic review [38] supported their use to optimize the knee and hip extension during the stance phase, reducing the metabolic cost of gait and increasing knee flexor and extensor muscle activity in children with CP. It seems that wearable exoskeleton devices could facilitate the activation of trunk muscles to guarantee an appropriate body alignment during weight-shifting movement, improving both static and dynamic balance, as well as gait [39]. In this vein, Kawasaki S. et al. [40] revealed that RAGT using an exoskeleton increased hip joint angles on the affected limb side promoting gait symmetry. In this way, patients were able to produce a better-grounded propulsion force, which is strictly correlated with plantar flexor strength. Interestingly, some authors [41] found that even a low dose of RAGT, using the Lokomat, could improve hip flexors and knee extensors muscle strength in moderate to severe CP patients.