In recent years, "exoskeleton robots" and "robotic legs" have appeared often in stroke rehabilitation advertisements and news reports. The pictures show patients wearing a mechanical device and walking forward step by step, like a scene from a science fiction film. Families naturally ask: can these devices help the patient walk again, and are they worth paying for? Robotic training is mostly self-paid and each session is costly, so families often have to make a decision in front of a large medical expense.
This article gives an objective analysis. It does not promote any equipment and does not follow trends. It uses the latest reliable clinical evidence to answer three questions: does robotic gait training or exoskeleton therapy help a patient walk again after stroke, which patients benefit most, and how cost-effective is it? In short, these devices are an add-on to conventional physiotherapy and do not replace it. They suit patients with specific conditions, and they are not worth using for every patient.
Summary: Using electromechanical-assisted gait training together with conventional physiotherapy (and not in place of it) can increase the chance that a stroke patient walks independently again. It has almost no practical effect on walking speed. The group that benefits most, and gets the best value, is patients with more severe stroke who cannot yet walk in the first 3 months after onset.
Clinical research results on robot-assisted gait training
The most reliable way to judge whether a treatment is worth using is to look at a systematic review that pools several randomised controlled trials. The Cochrane systematic review updated in 2025 (Mehrholz et al. 2025) examined the overall effect of electromechanical or robotic gait training combined with physiotherapy. Its conclusions are clear and also cautious:
The main benefit to expect from robotic training is that it helps the patient to cross the threshold from being unable to walk to being able to walk, and so raises the chance of regaining independent walking. The current evidence does not support an expectation that walking speed will increase greatly. The Cochrane report also warns that the benefit may fade after the course ends and may not be maintained in the long term. It is therefore not a one-time cure. It is an aid for increasing training dose during the main recovery period.
Limits of the effect: A gain of 0.05 m/s in walking speed is almost impossible to feel in daily walking. The main use of robotic training is to help patients who cannot yet walk to take their first steps again. It is not meant to make people who can already walk go faster. Understanding this helps the family use resources according to the rehabilitation goal and set reasonable expectations.
Which patients benefit most from robotic training
Beyond the average effect for the whole group, subgroup analyses that show which patients gain most are more useful. An earlier Cochrane systematic review (Mehrholz et al. 2017) had already shown that the benefit is not evenly spread across all patients:
For chronic-stage patients who can already walk slowly, the current evidence does not show that robotic training is better than conventional training. For the neuroscience behind "most benefit in the first 3 months" (sensitive periods and neuroplasticity), see the page on the stroke recovery "golden period". For the stages of the whole recovery process, see the stroke recovery timeline. The best time to start robotic training matches the acute and subacute phases.
Patients can also be grouped by the degree of lower-limb impairment. Patients with more severe stroke who cannot yet walk and need a lot of help can read physiotherapy for the lower limb after severe stroke. Patients who can stand but still have difficulty walking can read physiotherapy for the lower limb after moderate stroke, which has specific training programmes for each level of severity.
Cost-effectiveness of robotic training
Robotic training is mostly self-paid and each session is expensive, so whether it is worth joining depends to a large extent on cost-effectiveness. A cost-effectiveness analysis in Singapore (Shankar et al. 2025) compared a robotic exoskeleton with conventional physiotherapy in inpatient rehabilitation patients in the acute to subacute phase (under 9 months). It also showed that patients should be sorted by their condition, and that the device should not be used for everyone:
Based on the clinical evidence above, patients and families can use the following guide for a first assessment:
- Acute or subacute phase in the first 3 months after onset
- Patients with more severe stroke who cannot yet walk (FAC 0) and need a lot of help
- Hospital or institutional rehabilitation, where a full course can be completed
- The goal is to take the first independent steps again
- Chronic-stage patients who can already walk slowly
- Mild stroke with walking that is already largely independent
- The only aim is to walk faster, which the evidence does not support
- Putting resources into high-dose overground and functional training may be more practical
Choose the type of training according to actual need.If the patient's condition is mild, is already in the chronic stage, or the patient can already walk, spending the same resources on high-volume overground training close to daily life is usually better value. Whether a patient suits robotic training should be assessed by a physiotherapist according to severity, stage of recovery and goals, and the decision should not be based on advertising alone.
Comparing the two types of robotic equipment
Robotic walking systems on the market fall broadly into two types, and each has a different evidence base. The comparison below is objective and does not promote any particular model:
End-effector type (footplate type)
- The patient stands with both feet on moving footplates, and the machine moves the feet in a walking pattern.
- In subacute patients, this type of device gives a clearer improvement in walking speed.
- Even if the patient cannot yet control the legs, the machine can move them through a large number of repeated cycles.
Exoskeleton type (wearable)
- This is a mechanical frame worn on the legs. It follows each joint and gives powered assistance.
- Overground exoskeletons mainly help walking speed. The improvement is small but statistically significant.
- They usually give no clear extra benefit for balance (BBS), the Timed Up and Go test (TUG) or endurance (Lee et al. 2024).
The systematic review of overground exoskeletons (Lee et al. 2024) also found that these devices give limited extra help for balance and endurance, so one machine should not be expected to solve every functional problem. Improving balance and preventing falls needs a separate, specific programme, described in balance and fall-prevention training. Applying the regained walking ability to daily life (for example, arranging toilet use or going shopping) is the area of occupational therapy. Physiotherapy (PT) focuses on restoring the walking movement, and occupational therapy (OT) focuses on practical application in daily life. See occupational therapy after stroke for details.
Timing and frequency of robotic training
Another common misunderstanding is that "the more robotic training, the better". Evidence on overground exoskeletons (Lee et al. 2024) shows that the training parameters have an optimal balance, and that simply adding sessions does not produce a better effect:
- Each session lasts longer than 30 minutes, to give enough practice volume.
- No more than 3 sessions a week, so the body has time to rest and adapt.
- A course of about 4 weeks, concentrated in the effective rehabilitation window.
- Do not simply add more sessions in the belief that more frequent training is always better.
The machine is only an aid for increasing training dose. What actually drives functional recovery is high-dose, task-based repetitive training, together with the physiotherapist's assessment of the individual patient and programme planning. Home self-help rehabilitation equipment is also at a different level from the clinic or institutional robotic gait systems. For equipment suited to use at home, see stroke rehabilitation equipment and aids.
Use resources according to the patient's situation.Robotic gait training is not a miracle treatment and cannot replace the physiotherapist's professional judgement. Its role is to help patients with more severe stroke who cannot yet walk to take steps again during the main recovery period, and to increase their practice volume. Because it helps mainly in particular situations, it gives the best value when used at the right time and with the right goal.