Four clinical gait systems, four different architectures. A specification-led comparison of workflow, assistance and data transparency.
ExoSelect researchShowing English original
Continuer la lecture
Quick answer. Four clinical gait systems, four different architectures. A specification-led comparison of workflow, assistance and data transparency.
Lower-limb rehabilitation exoskeletons are often compared as if they were one product category. They are not. The four systems in this analysis differ in balance strategy, control model, portability and intended setting. The right choice depends on the clinical pathway, not a single specification winner.
EksoNR is a therapist-supervised gait-training system with a mature clinical workflow and software layer. Indego is modular and can be configured for clinical or personal mobility use. HAL uses bioelectric signal control and is often delivered through treatment or lease programmes. Atalante X is self-balancing and designed for overground training without parallel bars.
EksoNR and HAL assume a therapist is actively setting or supervising assistance. Atalante X is designed to manage balance itself, changing the therapist's role toward cueing and movement quality. Indego's modularity changes the logistics of fitting and transport.
These differences matter more than a few newton-metres of torque. A department with limited therapist capacity may value self-balancing; a department with a strong gait-lab workflow may value fine-grained assistance control.
All four systems disclose less than a complete engineering datasheet. Weight and battery runtime are generally available; peak torque, control-loop details and service-life data are often not. In the ExoSelect index, those gaps are explicitly marked rather than filled.
| System | Balance | Control model | Typical setting |
|---|---|---|---|
| EksoNR | Therapist-supported | Variable assistance | Inpatient / outpatient rehab |
| Indego | Therapist-supported | Intuitive / therapist | Clinical and personal mobility |
| HAL | Therapist-supported | Bioelectric signal | Neurorehabilitation programme |
| Atalante X | Self-balancing | Robot-managed gait | Overground clinical training |
No public specification can quantify patient comfort, donning time with a particular body type, or the quality of local service. Those require a structured evaluation with your own clinicians and patients.
Use the side-by-side tool to isolate the parameters that matter to your pathway, then request the missing values in writing. A maker that can supply a source for each number is a better partner than one with a prettier brochure.
An exoskeleton analysis is only useful if it changes what a buyer does next. The purpose of a dossier is not to declare a winner in the abstract, but to make the next conversation more specific: which task, which users, which environment, which constraints and which evidence is still missing.
The first step is to translate the task into a written requirement. Describe the posture, load, repetition, duration, reach height, floor surface and clothing. Add the practical constraints that decide adoption: donning time, heat, tool clearance, cleaning and storage. A system that performs well in a demonstration but interferes with the actual workflow will not be worn after the pilot team leaves.
The second step is to separate the system from the deployment. A device can be well engineered and still fail because the service network is thin, the training is inadequate or the spare-parts lead time is measured in months. Ask for the service model in writing, including response time, replacement units and the cost of consumables.
Data quality should be scored explicitly. A dossier that publishes weight, runtime, user limits and certification scope is easier to pilot than one that leaves those values unknown. Missing values are not automatically disqualifying, but they are a cost: someone has to resolve them before a purchase order, and that work belongs in the project plan.
Use the product index to narrow the field by category, drive type, weight, runtime and region. Treat the filters as a way to structure the question, not as a substitute for reading the parameter states. A system that disappears from a weight band because its weight is not disclosed is telling you something about the dossier.
Place the shortlisted systems side by side in the comparison tool and focus on the rows where they diverge. Differences are useful; similarities are useful too. If two systems are identical on the parameters that matter, the decision may turn on service, fit or commercial terms rather than engineering.
Check the manufacturer profile for verification level, product line and company documentation. A maker with a clear primary source for each claim is easier to audit than one whose brochure is the only available reference. Where a claim matters to safety, ask for the declaration of conformity and the technical file summary.
Finally, write the brief. Include the task, environment, user profile, deployment volume, region and budget range. A good brief lets the research desk separate sourced values from open questions, so the next conversation starts with evidence rather than a demonstration.
Training is part of the system, not an afterthought. Ask who delivers the first session, how competency is assessed and what happens when a new operator joins the team. A device that requires a specialist to fit it every morning has a different operating cost from one that a trained supervisor can deploy independently.
Keep a decision log beside the shortlist. Record what was verified, what was assumed and who owns each open question. When the recommendation goes to a capital committee, the log is often more persuasive than another product demonstration because it shows exactly which risks have been retired.
Send the requirement when the shortlist is ready. The response should make the gaps visible, not hide them behind a polished recommendation. That is the standard this index is built to support.
Start with the intended use, the certification scope and the service model. Those three items decide whether a system can be deployed, not just demonstrated.
No. Published parameters support shortlisting, but fit, comfort, training quality and local service response require a structured evaluation with the intended users.
Missing values are recorded as Unknown or Not disclosed and are never inferred from a similar product. Each dossier shows the date of the last update and links to its primary sources.
Editorial note
Cette analyse repose sur des spécifications publiées et des sources nommées. Aucune évaluation physique directe n'a été réalisée.
Relu par la rédaction ExoSelect
