dorsaVi has unveiled Sensor v6.5, its first wearable sensor that processes and acts on movement data in real time without cloud reliance, marking a key milestone in wearable Physical AI technology.
- Sensor v6.5 runs AI directly on the wearable device
- Built on FDA-cleared, TGA-certified platform with proven clinical validation
- Future roadmap includes RRAM and neuromorphic processing for ultra-low power
- Targets healthcare, elite sport, defence, and industrial applications
- Enables configurable biofeedback within milliseconds of movement events
Breaking the Cloud Dependency in Wearable AI
dorsaVi Ltd (ASX:DVL) has taken a significant leap in wearable technology with the launch of Sensor v6.5, an intelligent device that processes human movement and delivers feedback in real time; directly on the sensor. Unlike conventional wearables that rely on cloud or base device connections, Sensor v6.5 completes the full measure-analyse-decide-act loop on the device itself, eliminating latency and connectivity constraints.
This innovation positions dorsaVi at the forefront of Physical AI, a category of artificial intelligence that senses and acts within the physical world instantaneously. While Physical AI has been making strides in robotics and autonomous vehicles, achieving this capability on a body-worn sensor has been a tougher challenge due to power and compute limitations. Sensor v6.5 is dorsaVi’s first commercial product to overcome these hurdles, running on conventional silicon but designed with a clear roadmap to integrate advanced memory technologies like Resistive Random-Access Memory (RRAM).
Clinical Validation Meets Cutting-Edge Hardware
Built on a decade-old FDA-cleared and TGA-certified platform, Sensor v6.5 leverages a clinically validated principle proven in a 2015 randomised controlled trial: delivering biofeedback at the moment a movement fault occurs significantly improves patient outcomes. The device can distinguish normal movement, elite athletic performance, and injury-risk patterns, providing configurable haptic, audible, or visual feedback within milliseconds.
This real-time biofeedback capability is particularly valuable in healthcare and rehabilitation, where immediate corrective signals can reduce re-injury risk and enhance recovery adherence. The technology is also primed for elite sports, defence, and physically demanding industries, where monitoring fatigue and movement quality can prevent lost operational time and injury.
Advancing Towards Always-On Ultra-Edge Intelligence
dorsaVi’s commercialisation strategy extends beyond Sensor v6.5 with its Ultra-Edge Modular Design and Build program, which aims to embed next-generation memory and compute technologies directly into future sensors. The roadmap focuses on RRAM for low-power, non-volatile memory; compute-in-memory architectures that eliminate energy-wasting data shuttling; and neuromorphic processing that mimics the nervous system’s energy-efficient response patterns.
These innovations target the power envelope of body-worn devices, aspiring to enable always-on operation on coin-cell batteries; a feat conventional silicon architectures cannot achieve. The company’s recent progress on its 22-nm RRAM-CMOS validation chip underscores this trajectory towards scalable, energy-efficient AI hardware designed specifically for wearables and edge applications.
Expanding Applications Beyond Human Movement
While Sensor v6.5 currently focuses on human biomechanics, dorsaVi envisions its sense-decide-act core as a foundation for broader Physical AI markets. By pairing the sensor’s intelligence with machine-control interfaces, the platform could extend into exoskeleton control, collaborative robotics, and industrial automation without reinventing the sensing or decision-making components.
CEO Mathew Regan emphasised this dual strategy: “The sensor and the custom hardware we are developing are one strategy, not two: on-body Physical AI ready for market now, and the engine to power it more efficiently as it matures.” This approach leverages dorsaVi’s regulated, clinically proven platform as a springboard into emerging AI-driven industries.
Bottom Line?
Sensor v6.5 marks a tangible step from wearable measurement to embedded intelligence, but the pace and scale of commercial adoption and integration of advanced memory tech will be critical to watch.
Questions in the middle?
- How quickly will Sensor v6.5 gain traction across healthcare, sport, and defence sectors?
- What timelines can dorsaVi set for integrating RRAM and neuromorphic chips into commercial products?
- How will competitive pressures shape the adoption of on-device Physical AI wearables?