Bota Systems Cutting-Edge Force Torque Sensors Reviews

• 2021

Using the MiniONE Pro robotic arm in conjunction with ROS (Robot Operating System) for a variety of applications. I am impressed with the efficiency and ease of use of the MiniONE Pro, as well as its compatibility with ROS, which has allowed me to streamline the programming process for complex tasks.

One of the main advantages is its compatibility with the Robot Operating System (ROS), which has made integrating it into my robotic systems seamless and straightforward.

Lack of comprehensive documentation for the firmware side can be difficult. I also had some issues with the durability of the cable connector, as it seemed prone to breaking. While FT calibration and gravity compensation are available out of the box, inertial compensation is still missing and would be a helpful feature to add.

• 2021

I used this sensor in a custom-built testbed for measuring wheel-soil interaction forces.

The sensor was easy to mount and use, and support from the team was excellent when I had software issues. It was easy to use the sensor with ROS, and the data quality was very good.

The sensor takes 30-45 minutes to warm up to a steady temperature, and force measurements drift with temp, so you have to turn on the machine well in advance of data collection. The cable that came with our sensor was also fragile and needed to have strain relief added.

• 2022

LBR iiwa 14 R820 KUKA + BFT-MEDS-SER-M8 (Medusa 6-axis FT sensor with side serial interface)

Application: autonomously guide an ultrasound probe connected at the end-effector of a robotic manipulator. In order to guarantee continuous ultrasound feedback I required to ensure contact between the ultrasound probe and the patient skin. At first I used the estimated force/torque feedback provided by the KUKA arm, but then I preferred to move to a FT sensor placed before the ultrasound probe so to have an accurate and configuration independent force/torque feedback. This also allowed us to implement a force closed loop for autonomous scanning that is independent from the robotic manipulator adopted.

The sensor is compact, light, has a wider range (and a higher noise free resolution) along the z axis and it's very accurate. All the filtering is already performed on the sensor side and the signal obtained is ready to be used in the user application. It is easy to setup and it's directly compatible with ROS. I would add that customer service has been always very friendly, reactive and effective.

Absence of documentation particularly for the firmware side. I needed to have a call with the developers to understand how to properly configure the sensor for my application (and it is still a little obscure to me). It would be really helpful to have more information. Similarly, FT calibration, gravity compensation are out of the box but requires some work to be properly put in place. Inertial compensation was missing at the time I integrated the sensor in the system. It would be a nice feature to add if still not present.

• 2022

ANYmal from ANYbotics which we modified with an arm on top

I was implementing a teleoperation algorithm with force feedback for teleoperating a robot arm on a quadroped robot. That was part of the ANA Avatar Xprize challenge. The force feedback from joint torques was noisy and we needed a force sensor which is solid enough to handle harsh hits and accurate enough at the same time.

The sensor is extremely robust. It handles harsh falls (quadrupeds fall badly and ours is 60 Kg) and still works without issues after re-calibration. It is built to last even in rough environments. I rarely had issues with it even when operating it for several hours in the sun.

A rugged protection for the cable connector could help especially for applications like mine. It is the only thing which could break in the sensor.

• 2022

SCALER: Spine-enhanced Climbing Autonomous Limbed Exploration Robot, from UCLA

We used this sensor in order to achieve force control for the purpose of free-climbing. Our robot consists of multiple position-controlled motors and to achieve closed-loop force control, we need to have reliable force / torque sensors. However, as far as I know, there are not so many force / torque sensors which are so rubust. I know there are great sensors whose sensor accuracy is super good. However, if you simply attach those sensors to the end-effector of legged robots, they can easily be broken. However, the force torque sensors from BOTA are very robust. We did a number of experiments (more than in total 50 hours I guess) with trot gait of 8 kg robots but still they are not broken. Also the accuracy is very reasonable, which enables us to design good state and contact estimator for legged robots. Using BOTA FT sensors, our robot is now able to achieve closed-loop control with force and torque feedback. I really enjoy using them.

I would say the toughness. This sensor is so tough so that under various impulse experiments, the sensor has never been broken. Also, sometime, we accidently drop the sensors but still they are quite tough. I also have never seen inaccurate measurements of the sensors. If you are interested in using force torque sensors for tough condition, I think this is the right sensor you should consider.

Even though they have a great ROS package for running the sensors, sometimes we want to use the sensors directly by compiling the source. However, unfortunately, it never works. Also, they have public codes on their gitlab but it is a bit difficult to understand what is going on since there are not many explanations.