An FSR sample should not be approved from one resistance-versus-force curve. Approve the complete input chain: the real actuator, its contact area, preload, support surface, electrical circuit and the software rule that turns the signal into a command. If any one of those changes after sampling, the same sensor can feel slower, saturate earlier or generate a less stable command. Freeze the mechanical and signal-mapping conditions before calling the sample representative.
What the August 22 cursor-control study changes
Combining force and kinematic signals for human-computer interface: insights from cursor control by flexible pressure sensors was published online on August 22, 2026 in Ergonomics. The original experiments compared mappings based on center-of-pressure movement, force and a hybrid of the two for cursor control in space-constrained conditions. The hybrid mapping had the highest throughput in the study, while the comparison also exposed a speed-versus-stability tradeoff.
The researchers did not test a Baoshengda FSR, and their mapping is not a ready-made specification for another product. The useful sourcing lesson is narrower. A pressure signal only becomes a usable command after the buyer defines how position, force and time are interpreted. A sensor supplier cannot infer that rule from the active-area outline alone.
Freeze the actuator before comparing sensor samples
The actuator is part of the sensing system. Record its diameter, edge radius, hardness, surface finish and alignment to the active area. A small hard tip can create a high local stress and an early output change. A larger compliant pad spreads the load and can delay or smooth the response. Neither result is inherently correct unless it represents the installed mechanism.
Preload also matters. A foam layer, spring, cover film or enclosure rib can hold the sensor above zero load before the user touches the control. That starting point consumes part of the usable range. Test the production-intent stack, not only a loose sensor under a laboratory weight.
Use the FSR pressure sensor engineering route to align the active area, tail direction and actuator drawing before electrical limits are frozen.
Link every command state to measurable evidence
A sample plan should connect the user's action to a measured signal and then to the host command. The table below is a starting structure, not a universal acceptance standard.
| Input condition | Mechanical evidence | Electrical or command evidence |
|---|---|---|
| No intentional input | Installed preload and support position | Baseline band with no false command |
| First detectable touch | Actuator contact location and travel | Named threshold and debounce result |
| Normal command force | Defined force span and hold time | Stable command with the required response |
| Off-center input | Offset from the active-area datum | Allowed output change or blocked command |
| Maximum expected input | Mechanical stop and overload path | No unintended adjacent command or permanent shift |
| Repeated movement | Cycle count and return interval | Drift, hysteresis and missed-input record |
Do not substitute a published force range from another sensor. The host circuit, divider resistance, ADC range, filtering and sampling rate can change the apparent response even when the mechanical sample is unchanged.
Run the sample in the mounted control sequence
Start with a baseline after the sensor has been assembled into the intended fixture. Apply the minimum, normal and maximum defined inputs through the real actuator. Hold each load long enough to observe settling and release it long enough to observe recovery. Repeat the sequence at the locations the user can actually reach.
For a position-sensitive command, move the actuator across the permitted travel and record both the location signal and force. For a single command, challenge the edge of the active area and nearby inactive regions. The test should reveal whether mechanical tolerances can create a false command, not merely whether the center responds.
Retest after tail bending, connector mating or soldering, depending on the termination. A sample can pass on a flat table and shift after the tail pulls on the sensing area or the enclosure compresses the stack. Keep the pre-test and post-test curves with the fixture revision.
Supplier boundary and limitations
Baoshengda can review the FSR active area, printed sensing pattern, spacer or stack concept, tail geometry, contact termination and physical sample evidence for an agreed design. The buyer remains responsible for the actuator, host circuit, firmware mapping, command safety, mounted mechanics, environmental limits and final equipment validation.
The cited study is current original research, not a Baoshengda test or a guarantee of cursor performance. Its throughput and trajectory results belong to the reported experimental setup. They should not be copied into an RFQ as a product claim.
FSR sample-approval package
Send the following items before requesting a production-intent sample:
- active-area outline with datum and permitted input locations;
- actuator diameter, hardness, edge shape, travel and alignment tolerance;
- support surface, cover layer, foam, spring and preload condition;
- minimum, normal and maximum input force with hold time;
- divider, amplifier or ADC circuit and supply voltage;
- threshold, filtering, sampling and command-mapping rule;
- allowed hysteresis, drift, repeatability and false-input result;
- tail exit, length, contact finish, connector or soldering method;
- mounted fixture and step-by-step sample test;
- prototype quantity, production estimate and revision identifiers.
When those fields are controlled, send the FSR drawing and sample conditions for quotation. The quotation can then cover a sensor and sample plan tied to the real actuator instead of a generic pressure curve.
Need help reviewing a structure?
Send your drawing, photos, application, and quantity. Baoshengda can help check the structure before sampling.
Send Drawing for Quote