Force/Torque (F/T) sensors are critical components of robotic applications in electrical and mechanical assembly, product testing, material handling, and many other applications. A sensor system helps robots verify part insertion; hold constant force during buffing, polishing, and deburring; and collect force information for lot testing and statistical process control (SPC).
The robotic sensor system consists of a transducer mounted on the robot and a sensor interface controller connected to the transducer by a high-flex cable. The transducer converts force and torque loading into strain-gage signals and transmits them to the sensor controller. The sensor controller collects transducer strain gage vectors, performs computations, and outputs F/T data directly to the robot.
Depending on the user’s requirements, selecting the correct sensor system may be a challenge. Thus, the most important activity to complete before selecting the actual components of a sensor system is identifying the application. Questions that need to be answered are:
- How will the robot be used?
- What range of force will it experience?
- What are the environmental conditions?
Once the application is clearly identified, focus shifts to the components that will make up the sensor system, namely transducers and sensor controllers.
The first step in selecting a transducer is to calculate the expected moment and forces. Moment capacity is usually the determining factor when choosing the best transducer for a robotic application.
The end effector, or mechanical tool, attached to the transducer will generate forces as it performs its task. The distance of the applied force from the transducer will result in a moment, or the applied force multiplied by the distance from the transducer origin to the point at which the force is applied (See Figure 1). It is important to consider overload conditions as well as normal operating forces and moments the transducer will experience.
Be sure to add in all loads the transducer will experience when calculating load on the transducer, including those the application does not monitor. Also be aware that the published payloads of robots are typically the maximum loads the robot can handle at published positional resolution. A robot can actually handle and create much larger loads, but with some loss of positional repeatability.
Robots are typically overpowered for an application and the robot is capable of exerting loads many times its rated loading. During a crash, for example, the inertia of the sudden deceleration can generate large loads and force of impact. A robot can generate deceleration of 5gs during an E-Stop (Emergency Stop) event.
The strain gage sensing technology used can also influence the transducer’s factor of safety. Transducers using high output strain gages can be designed to withstand higher overload conditions than designs using lower output strain gages. High output strain gages can also have lower noise levels since they require less signal amplification. Silicon strain-gages provide a signal 75 times stronger than conventional foil gages.
Next, transducer capacity needs to be determined. Specific information required to select the correct transducer model and calibration include: minimum and maximum Forces (Fx, Fy, Fz), minimum and maximum torques (Tx, Ty, Tz), weight, diameter, and height. Typically, sensor manufacturers provide a selection table that cross-references measurement ranges with types of transducers available (See Figure 2).
The third step is to verify the resolution and accuracy requirements of the application verses those of the transducer under consideration. A fine resolution requirement can conflict with a transducer chosen based on moment capacity. Transducers with larger ranges have coarser resolutions. The output resolution of a transducer is much finer than its absolute accuracy — be sure the absolute accuracy fits the application. Like single-axis load cells, six degree of freedom transducers have their absolute accuracy expressed as a percentage of their full-scale load for each axis.
A specific transducer should be apparent after completing the previous steps. Compare the detailed description of the transducer with application parameters to optimize performance. For example, when the expected maximum measured load is 55 pounds of force and the end effector is 8" long, the moment generated is 440 in-lbs; the best transducer using the table in Figure 2 would be Model E, which can handle Txy moments up to 600 in-lbs. It is also important to identify likely situations that could overload the transducer and verify that the transducer will not be damaged during these overloads.
Sensor controllers receive information from the transducer and produce resolved force and torque data. Onboard software calculates the output data by multiplying the strain-gage vector by a calibration matrix to form the F/T data consisting of three orthogonal forces (Fx, Fy, Fz) and torques (Tx, Ty, Tz) (See Figure 3). The force and torque data can be transmitted to the robot and serve as signals needed for the robot to perform the intended function.
Most commercially available sensor controllers provide the following functionality:
- Outputs all six axes of load data (Fx, Fy, Fz, Tx, Ty, Tz).
- Tool transformations allow movement of the center of origin to a user-specified location.
- Peak analysis allows minimum and maximum F/T values to be detected and stored.
- Biasing provides a convenient way to subtract unwanted loads from readings.
- Data filtering allows the user to minimize the effects of unwanted vibrations in the system.
- Programmable threshold monitoring with optically-isolated I/O connections provide high-speed response to the robots discrete I/O panel.
Sensor controller selection is determined by the output resolution of the sensor controller, the output format of the sensor controller, and available software to interface with the sensor controller. Some sensor controllers provide better resolution and noise performance than others. Commonly available output formats are: RS-232 (serial), analog voltage, and computer-bus (ISA, PCI, etc.). Check with the manufacturer to see if interface software is available to ease system integration.
Two fundamental types of sensor controllers are available: stand-alone and computer bus. The advantage of the stand-alone sensor controller is that it is self-powered and self-contained. They typically communicate with the robot controller via RS-232 serial format and/or via analog voltages. The sensor controller’s discrete I/O connections allow for easy connection to PLCs and other industrial equipment.
The computer bus sensor controller is targeted to a specific type of computer backplane and will plug into the robot or computer’s motherboard. Communication is achieved through software drivers such as ActiveX for Windows platforms, or directly to its I/O mapped registers. Since the computer bus sensor controller can be placed inside of the robot system it has a much cleaner appearance than the stand-alone type.
Software provided by the sensor controller manufacturer can often display the F/T information for all six degrees of freedom simultaneously on the computer screen, allowing the user to easily modify different measurement parameters and determine the current loading.
The type of sensor controller selected is often dependent on how the F/T information will be used. Sensor force and torque data can be used in several ways, such as data collection/analysis, real-time force control, and threshold detection. A quick review of types of data usage and corresponding sensor controller will help finalize the selection process.
For data collection, the computer bus sensor controller provides the easiest integration for PC users. Installed in the PC, it communicates directly with standard operating applications, such as LabView and Visual Basic. Data collection speeds can be influenced by computer speed and the Windows operating system.
Real-time force control is attainable using the ISA bus sensor by simply integrating with software drivers. All F/T data is available on the computer bus, allowing control software instant access. If users are not working in a PC environment, analog outputs created by stand-alone sensor controllers can be interfaced to any analog input card.
Force and torque threshold or limit detection is available on some types of sensor controllers. This capability allows the sensor controller itself to monitor transducer loads for specific loading conditions and notify the robot controller when the conditions have been met. By moving this monitoring function to the sensor controller the robot controller is relieved of the monitoring task. One example of this is to use the sensor controller to monitor for dangerous loads. When a dangerous load condition is detected the sensor controller’s discrete output triggers the robot’s E-Stop circuit.
The selected transducer must be electrically connected to its sensor controller. The manufacturer usually provides a standard length cable assembly. The user, however, must determine how much cable their application actually requires.
Keep in mind that the cable needs to reach from the sensor controller to the transducer in any robot position.
As previously stated, F/T Sensor systems are used in a variety of applications. Most suppliers manufacture a rugged and extremely durable transducer, virtually eliminating this aspect from the selection process, but it should still be taken in to consideration. Physical attachments, mounting plates, and tool transformations are all factors that must be evaluated.
Mounting the transducer to the robot is facilitated by several options, such as quick disconnects and standard or customized interface plates. Additional transducer options include temperature operating ranges, multiple calibrations, and unique operating environment requirements — nuclear radiation tolerance and MRI resistance for example. Custom designed and built sensor models may be needed to meet specific applications.
This article was written by Dwayne Perry, PE, chief sensor technologist for ATI Industrial Automation in Apex, NC. For more information, contact the author at