A walk down a corridor twice a year is not how anyone walks.
Why the measurement has to leave the clinic, in three findings from the literature. Then what we have done about it.
The case
Three things the field has already shown.
-
Clinic tests do not describe how people walk.
Performance tests in the clinic correlate only low-to-moderately with real-world walking speed. The two- and six-minute walks hold up best, which is why those are the two tests eSteps runs.
PLOS One
-
Daily-life gait shows change before the clinic does.
Stride-level measurement in daily life detects disability progression in multiple sclerosis earlier than the clinical scores in use today.
Lancet eClinicalMedicine, 2026
-
Regulators are moving the same way.
The Mobilise-D consortium, with EMA support, is qualifying real-world walking speed as a mobility endpoint for drug trials.
Digital Biomarkers · Frontiers in Digital Health, 2026
Our part.
Three studies against laboratory references and one pilot in patients, all at the University of Bologna and Policlinico Sant'Orsola. The numbers, as measured.
-
Timing to lab standard
Stride time within 0.007 s of force plates, no significant difference. Step length within 13 cm and speed within 0.21 m/s: good for following one person over months, not for replacing a lab measurement on the day.
-
Reads the activity
Walking, fast walking, running, stairs up, stairs down, still: 91% accuracy, scored on people the model never trained on. Peer-reviewed.
-
Seen in MS
A feasibility pilot at Sant'Orsola: patients walked the two-minute test in the insoles, the parameters came out, and they showed what multiple sclerosis is known to do to a walk - slower, shorter steps, less regular timing. Validation against EDSS is still running.
The two validation studies are MSc theses and are being prepared for publication. eSteps is a monitoring instrument. It does not diagnose, stage disease or recommend treatment. The CE mark and FDA registration cover gait monitoring; interpretation belongs to the clinician.
Research and clinical collaborations
If you want the underlying data, the protocols or the analysis, write to us.