Design of a Wearable Device to Study Finger-Object Contact Timing during Prehension Zhaoxin Yin 1 , Yih-Choung Yu 1 , Luis Schettino 2 1 Dept. of Electrical and Computer Engineering and 2 Neuroscience Program Lafayette College, Easton, PA, USA Abstract — A wearable device was developed to study the human grasping behavior. This device includes a custom-made glove with conductors at finger tips to detect the finger contact signals, a simple electronic circuit to enhance the signal quality, and a real-time data acquisition system to synchronize the acquisition of the finger contact signals with obtaining the optical marker signals in a motion capture system. A preliminary study shows that the new device is capable of detecting the moments of finger- object contacts with better resolution of time than the existing motion capture camera system by three orders of magnitude. This new device will be intensively used in experiments to understand human grasping. I. INTRODUCTION Grasping models have been used to research human motor cognition since the 1980s. The movement involves transport and preshaping of the hand, contact with the object, finger force development, and lifting of the object. Motor control researchers express a difference of opinion with regards to the organization of fingers prior to contact with the target object. While Smeets et al. [1] maintains that the digits move independently when forming contact, Jeannerod et al. [2, 3] characterized the digits move relative to each other in forming the grip. In addition to finger coordination prior to object contact, the actual temporal organization of the digits of the grasping hand is also unknown. Reilmann et al. [4] stated that finger contact timing depends on the timing of finger-object force development, i.e. one applies force to an object (for lift) only after s/he first contacts it with the fingers. Thus, only by linking the preshaping and force development studies together—via the temporal organization of digit-object contact timings—can such questions of digit coordination be investigated further. The actual timing of finger contact during object acquisition has not been studied in detail. The main reason for this lacuna in our understanding has been technical limitations. While motion-tracking techniques can provide precise temporal and spatial information about limb and/or discrete position during movement, the characteristics of the human hand, with five fingers contacting the object at a very rapid sequence in a relatively small volume of space, make it extremely hard to determine the order and temporal spacing between contacts. One can only qualitatively describe finger contact sequencing derived from studying fingertip force development [4]. Unfortunately, in these studies the subjects are confined to grasping the target objects at pre-specified locations, thereby precluding more natural grasping motions. There are still many unanswered questions. For example, it is not known what the maximum contact timing latency between fingers that will still permit a stable grasp is. Also, the relationship between hand preshaping and contact timing is not known. Schettino et al. [6] suggested that one reason for hand preshaping as a response to object shape is to reduce discrete timing differences, thereby favoring stable grasps. To address some of the limitations stated above, a system that can more precisely determine the time at which each individual finger contacts the target object was designed and implemented. This paper presents the design of the wearable data acquisition system and some preliminary data analysis. II. SYSTEM DESCRIPTION A. Configuration of the System The entire system, illustrated in Fig. 1, consists of two major devices: a data acquisition system (PCI-6036E), and a motion capture system (NaturalPoint Inc.). Both components are synchronized by a real-time data acquisition software Wincon 5 (Quanser Inc.) along with Simulink (MathWorks), installed on a personal computer (PC1 in Fig. 1). The data acquired by the data acquisition system provide the time indications of each individual fingertip contacting a metal-shielded object (maintained at 9V) occurred during grasping. The data captured by the motion cameras system register the motion of the limb, hand, and fingers, which provide important information to understand their motion at the occurrence of a finger-object (F-O) contact. By synchronizing both devices, the exact location of the upper limb at any time prior to the contact and the contact timing can be identified from the data obtained from both devices. Device synchronization was carried out with a master-slave architecture, in which the data acquisition system acts as the master device sending an activation signal to coordinate the operation of the slave cameras. Fig. 1, Block diagram of the system 978-1-61284-8928-0/11/$26.00 ©2011 IEEE