Design parameters for robotic tactile sensors are engineered from the human sense of touch. Robotic tactile sensing has been linked with detecting and measuring forces in a specific area alone.
Cutaneous or tactile sensing is related to detecting and measuring contact parameters that include mechanical stimulation such as force, stress, roughness, moisture, or temperature.
Tactile sensing may be termed as the method of detecting and measuring a specified property of a contact event in a particular area and then pre-processing the signals at the sensor level before sending it to higher levels for perceptual interpretation.
The different types of tactile sensors used in robotic applications are:
- Resistive sensors
- Piezoresistive sensors
- Tunnel effect tactile sensors
- Capacitive sensors
- Optical sensors
- Ultrasonic sensors
- Magnetism-based sensors
- Piezoelectric sensors
- Piezoelectric polymer sensors
- MEMS sensors.
Dahiya R.S et al. from the Italian University of Technology and the University of Genova developed robots that were capable of operating in unstructured difficult environments or those that were suited to challenging conditions inaccessible or hazardous for humans and for that purpose they needed advanced sensory capabilities far beyond those normally available.
This study presented tactile element arrays that were placed on the distal phalange of the humanoid robot present in their lab. One design included 32 element microelectrode arrays having 1 mm centre-to-centre distance. Each tactile element or taxel includes a sensing material mounted on a microelectrode.
Each taxel will be used as an extended gate of a FET and the voltage or charge generated is collected by the taxel due to the applied stress. The FET devices and the taxels array are integrated on the same silicon die in the second design.
In June 2012, a study was published in Frontiers in Neurorobotics by Fishel J.A et al (2012) which demonstrated that it is possible for a specially engineered robot to surpass humans in the detection of a broad range of natural materials based on their textures, enabling advancements in prosthesis, consumer product testing and personal assistive robots.
A new kind of tactile sensor was integrated in the robot to simulate the fingertip of a human. The sensor is capable of a number of human sensations and can tell where and in which direction force is applied to the fingertip and information on the thermal characteristics of an object being touched.
The BioTac sensor is equipped with a flexible, soft skin on a liquid filling and its tactile sensor modalities are demonstrated in the video below.
Fingerprints are also present on the skin surface improving its sensitivity to vibration considerably. The skin vibrates in typical ways as the finger slides over a textured surface. A hydrophone identifies the vibrations inside the finger core, which is a bone-like structure. The BioTac is more sensitive than the human finger.
The robot, built by Fishel, underwent training on 117 commonly found materials gathered from stationery, hardware and fabric stores. The robot was able to rightly identify the material 95% of the time selecting intelligently and making an average of five exploratory movements. According to researchers, this touch technology will be useful as human prosthesis technology or for assisting companies that employ experts for judging the feel of consumer products and human skin.
Gwilliam J.C et al (2010) from John Hopkins University proposed tactile sensors that could identify lumps in robot-assisted palpation. The study compared the performance of a capacitive tactile sensor with a human finger. The sensor response was evaluated as it pertains to robot-assisted palpation and the senor performance was compared to human subjects doing a similar task on the same set of artificial tissue models.
Also, the impact of a number of tissue parameters on the performance of the tactile sensor and the human finger were evaluated. With the help of signal detection theory for determination of tactile sensor lump detection thresholds, the tactile sensor proved to be better than a human finger for a palpation task.
Three levels of difficulty have been identified for robotic tasks, which include irregularities in the handled objects, disorder in the working environment and both these combined together. Some examples are food processing in an industrial set-up, underwater repair and surgeries, respectively. As touch and contact are highly significant, tactile sensing will offer solutions unmatched by sensing modalities like vision.
Hence tactile sensing finds applications in:
- Surgical and medical robots – In surgery tactile sensing is of primary importance. Tactile sensing and telepresence are especially important for laparoscopy and keyhole robotic surgery to overcome the problems of lack of depth from 2D vision, restricted manipulation mobility and almost no sense of touch.
- Health care and service robots – The key issue here is that the robots will have to work in a human-centered environment that contains irregular and unstructured objects of different composition and structure. Hence there is a need for tactile sensing as well as dexterous manipulators to retrieve items for the aged and disabled.
- Agriculture and food processing – Robots used in these sectors will need robotic handling systems that have good adaptation, high sensitivity and dexterity. In order to manipulate, grasp or process soft objects like confectionery, fruit and food items, there is a need for the sensor to sense textures, hardness and surface properties.
Tactile sensing complements vision. It is important to understand that vision and other sensors involve sensing from a distance whereas this involves direct contact. Over the years, advancements with robotic hands having a number of fingers and hapti sensing has shown great promise.
Laboratory systems are now able to maintain control over objects whereas during manipulation tasks, rolling and sliding occurs. It is anticipated that there will be a great amount of development in medicine and surgery.
The dependence of surgeons on tactile feedback has inspired research in restoring tactile sensing in MIS and other techniques. Remote palpation and novel sensors for medical tactile exploration are other medical examples.
Sources and Further Reading
- Fisher J.A and Loeb G.E. Bayesian exploration for intelligent identification of textures. Frontiers in Neurorobotics. 2012; 6:4. doi: 10.3389/fnbot.2012.00004
- Massaro A, et al. Improvements of optical tactile sensors for robotics system by gold nanocomposite material. Journal of nanoscience and nanotechnology. 2012;12(6):4878-82.
- Dahiya R.S and Valle M. Tactile Sensing for Robotic Applications. Sensors, Focus on Tactile, Force and Stress Sensors, Book edited by: Jose Gerardo Rocha and Senentxu Lanceros-Mendez. Page 444, December 2008, I-Tech, Vienna, Austria.
- Gwilliam J.C, ET AL. Human vs. Robotic Tactile Sensing: Detecting Lumps in Soft Tissue. Published in Haptics Symposium, 2010 IEEE. 2010; 21-28.