A 25-50% sensor cut and sub-30ms latency target moves the company from measurement vendor to control-layer partner
dorsaVi (ASX:DVL) has put numbers around something investors have been waiting to see costed. The company’s RRAM and neuromorphic roadmap, applied to lower-limb exoskeletons, is projected to cut sensor count by 25 to 50%, reduce wireless data volume by 10 to 100 times, and more than double battery life.
The framing matters more than the percentages. dorsaVi is not pitching itself as an exoskeleton builder. It is pitching itself as the sensing and intelligence layer that sits inside someone else’s exoskeleton, telling the robotic frame what the wearer intends to do before the leg actually moves.
- Updated
- INSIDER CONVICTION TRADES FY26
The market hasn't noticed yet.
Disclosed by law. Missed by most investors. 129 trades tracked by us.
NO Credit card
- Director buys & sells tracked
- Buy / Hold / Sell verdict on every stock
- Technical analysis included
That is a deliberately narrower position than the broader robotics story, and a more commercial one. It makes dorsaVi a partner to exoskeleton manufacturers rather than a competitor, and it opens four revenue paths from sensor sales through to recurring software and validation services.
The global exoskeleton market is projected at US$1.79 billion by 2033, spanning medical rehabilitation, industrial safety, defence and aged care. dorsaVi’s pitch is that the bottleneck across all four verticals is the same one. The intelligence layer, not the hardware.
Why the in-sensor computing shift actually changes the economics
The current dorsaVi platform measures muscle activity using EMG (electromyography, the electrical signal muscles produce just before they contract) and streams the raw data to an external controller. That works for analytics and validation, but it burns battery, eats wireless bandwidth, and introduces latency the closed-loop control inside an exoskeleton cannot tolerate.
RRAM (resistive random-access memory, a fast low-power non-volatile chip technology) combined with neuromorphic processing (brain-inspired local computing) is what lets the sensor itself do the thinking. Instead of streaming raw EMG, the sensor transmits compact decisions. Swing intent probability, fatigue index, gait phase confidence.
The projected target performance is sensor-node decision latency below 10 to 30 milliseconds and real-time classification accuracy above 90%. If dorsaVi gets near those numbers, the sensor stops being a wearable measurement device and starts being a control node.
Existing FDA-cleared hardware already supports the near-term path
The part of this announcement that gets understated is that dorsaVi’s current EMG and movement sensors, already FDA-cleared and TGA-certified, can support five exoskeleton use cases today. Gait intent detection, gait phase recognition, assist-as-needed tuning, fatigue monitoring, and exoskeleton validation.
The deployment path is staged. Eight sensors for full bilateral lower-limb mapping in early work, converging to four to six sensors for real-time supervisory control, and simplifying to four EMG plus one motion sensor for field validation. That progression is what makes the commercial pitch credible.
We think this is the more important near-term story. Exoskeleton manufacturers running clinical or industrial trials need a validation partner now, not in two years when the RRAM silicon lands. dorsaVi can sell into that gap with hardware that already exists.
What investors should be skeptical about
The projected figures are projections, not validated benchmarks. The 2x battery improvement, the 25 to 50% sensor reduction, and the sub-30ms latency targets are derived from technical analysis rather than fielded results. The Ultra-Edge Modular Design and Build program announced 6 May 2026 is still translating IP into manufacturable hardware.
Our concern is that announcements of this kind, projecting performance gains from a roadmap rather than reporting them, are easy to write and harder to verify. The next test is partner engagement. Specifically, named exoskeleton OEMs running validation programs using dorsaVi sensors, with disclosed terms.
Until that arrives, the exoskeleton extension is a strategic positioning statement supported by an existing sensor business. That is not nothing, but it is also not yet commercial proof.
The Investors Takeaway for dorsaVi
dorsaVi has done the work of mapping its existing sensor platform onto a clearly defined commercial wedge in exoskeletons, with a staged technical path and a longer-term silicon story behind it. That is a more disciplined pitch than most ASX-listed neuromorphic stories investors have seen this cycle.
What we will be watching across the next two quarters is whether the company can convert the analysis into a named partner program. An exoskeleton OEM trial, a defence validation contract, or a rehabilitation robotics integration deal would each move the story from projected to proven. Without one, this remains a roadmap pitch with credible foundations rather than a revenue inflection.
Investors looking for more in-depth coverage of ASX-listed wearable and neuromorphic technology names can find it at stocksdownunder.
