Hi all, I’m Kasra Shariat (CEO, Cypress Innovative Solutions Ltd, UK). We’re developing a compact 6-axis force/torque sensor prototype for underwater robotics payload integration and test/verification rigs, where repeatable multi-axis force feedback helps with contact-rich tasks (docking/contact, tool engagement fixtures, and validation).
Key specs: 0.029 N force resolution and 2 mN·m torque resolution. Strong cost advantage vs comparable multi-axis sensors because we control design and production in-house. Customization: measurement ranges, mechanical interface/flange/stack-up, materials/packaging, comms (UART today; CAN/EtherCAT options). Working lab prototype with an industrialization roadmap.
If you’re interested in evaluating for your vehicle/tooling or test setup, reply here or DM me.
It would likely be interesting to the community here to know how involved an integration would likely be with standard Blue Robotics vehicles, and what kind of price range you’re targeting.
Given you’ve tagged this post with ardusub, does that mean there’s an ArduPilot driver available already for the sensor?
More generally, it doesn’t seem like the sensing range is provided in the specs - only the resolution, accuracy, and sample rate.
Integration with standard Blue Robotics vehicles:
The current prototype outputs calibrated F/T data over UART. On a typical BlueROV2/ArduSub stack, integration can be as simple as wiring UART to an onboard computer (or microcontroller) and logging/using the data at the application layer (e.g., Python/ROS/BlueOS container). Mechanical integration is via a compact flange/stack-up at the payload/tool interface.
ArduSub / ArduPilot driver:
Not yet as an upstream ArduPilot driver. I tagged ardusub mainly because the likely user base is BlueROV2/ArduSub builders. If there’s interest, we can provide (a) a serial protocol spec and reference parser, and (b) work toward an ArduPilot integration approach (e.g., a companion-side bridge first, then a proper driver if it proves useful to the community).
Price range:
We’re aiming for a material cost advantage vs comparable 6-axis sensors, but final pricing depends on the measurement range + packaging + interface requirements. If you share your target range/environment constraints, I can give a realistic ballpark for an evaluation unit.
Sensing range:
Good point. We intentionally didn’t lock a single range because ranges are configurable and there are trade-offs (range vs resolution/accuracy, mechanical stack-up, etc.). If you tell me the typical force/torque envelopes you expect for your use case (tooling contact, docking, sampling), I’ll share the candidate ranges we’re supporting in the current prototype and what’s feasible for a ruggedized version.
If anyone here has a common “standard” range they’d like to see for BlueROV2-class tooling, happy to align on that as a baseline.
