**1. Introduction**

24 Haptics Rendering and Applications

Streri, A. (1987). Tactile discrimination of shape and intermodal transfer in 2- to 3-month old

Volcic, R., Wijntjes, W. A., Kool, E. C., & Kappers, A. M. L. (2010). *Experimental Brain* 

Woodworth, R. S., & Schlosberg, H. (1954). *Experimental Psychology*. (2nd ed.). New York:

Zaki, S. R., & Homa, D. (1999). Concepts and transformational knowledge. *Cognitive* 

infants. *British Journal of Developmental Psychology, 2*, 287-294.

*Research, 203*, 621-627.

*Psychology, 39*, 69-115.

Holt, Rinehart & Winston.

The development of human/computer interfaces related to human haptics is subject to greater degrees of difficulty and complexity than those related to human visual and/or auditory senses. Such interfaces require direct contact and/or interaction with humans as opposed to visual and/or auditory devices that do not require such direct contact, manipulation and other force-related interactions. Therefore, without a deeper understanding of the mechanisms involved in haptics, such interfaces may be far from being user-centered or easy-to-use.

The focus of our research has been to understand the perceptual system of heaviness in humans. Heaviness perception, categorized as one aspect of haptic perception, is considered to be a vital ability in everyday life not only to recognize objects, but also to lift and manipulate them. Research in the field of experimental psychology, in particular psychophysics, has focused on identifying the properties and mechanisms of heaviness ever since Weber (1834, as translated by H. E. Ross & Murray, 1978) undertook his inquiries. Such properties and mechanisms have not yet been fully identified. Rather, the more experimental techniques and/or experimental environments have evolved, the more complex human perception of heaviness appears. This is because heaviness: (1) involves both perceptual systems and sensorimotor systems, such as the force programming system for lifting or holding objects, (2) is affected not only by object weight, but also by physical, functional and other properties of objects and (3) is affected by bottom-up processing by lower-order senses and by top-down processing by higher-order cognitive processes such as expectation and rationalization.

The purpose of this chapter is to overview human perception of heaviness to decipher its complexity. In addition, we introduce the usefulness of virtual reality systems to isolate and understand constraints on heaviness perception. One such system adopted for our research is the Virtual Hand Laboratory, creating virtual or augmented environments, in which humans interact with computer displayed objects or real physical objects. We illuminate mechanisms of heaviness perception with fundamental findings that might not have been

Computer Graphic and PHANToM Haptic Displays:

and a feedback system.

McCloskey, 1978).

object accurately.

Powerful Tools to Understand How Humans Perceive Heaviness 27

must precede these behaviors (MacKenzie & Iberall, 1994). When an object is lifted by an adult, the sensorimotor system functions automatically, without any particular attention, to achieve a safe and effective grip to lift and hold, transport or manipulate the object, maintaining stability. This system is composed of two subsystems: a feedforward system

A feedforward system works before the initiation of the grasping movement to predict object properties including weight and estimate the required motor commands, based on long-term memories (Gordon et al., 1993) and/or short-term memories if the lifts are repeated during a short period of time (Johansson, 1996). Finally, these programmed motor commands are sent to the related muscle groups to achieve stable lift of the target object. Interestingly, the motor commands are hypothesized to be sent not only to the effectors, i.e., related muscle groups for object lifts, but also for "efferent copies" to be sent to the CNS structures related to sensation, feedback processing or perception (Sperry, 1950; Holst, 1954;

Once the lift of an object has been initiated, the feedback system acts to optimize the output forces in the muscles through a local reflex feedback via Ia afferents in a muscle spindle (for muscle stretch) and Ib in a Golgi tendon organ (for muscle tension) (Crago et al., 1982; Rothwell, et al., 1982). These peripheral-origin signals ascend via the dorsal column through transcortical loops in the central nervous system (CNS) relating to motor control (S1, M1, basal ganglia, and cerebellum), and ongoing motor commands are probably modified via cortico-cerebellar connections with the CNS-origin signals "copied" to optimize subsequent discharge based on detected errors between the efferent copies and afferent information. The optimized signals, then, may contribute to achieving safer and more stable lift (Brodie & H.E. Ross, 1985). Grip forces applied at the object/digit interfaces are also automatically adjusted, based on the information from mechanoreceptors on glabrous skin, from the initiation of lifts according to frictional forces, object slipperiness (Johansson, 1996;

The "copies" have been termed in such various ways as "sense of effort" (McCloskey, 1978), "collorary discharge" (Sperry, 1950) or "efferent copy" (Holst, 1954). Furthermore, in psychological research such as that for the size-weight illusion (SWI), (Charpentier, 1891 as cited in Murray, et al., 1999; See Sec. 3), efferent copy is replaced by the term "expectation" (H.E. Ross, 1969) or "ease with which could be lifted" (Müller & Schumann, 1889, as cited in Davis, 1973). These copies are thought to play an important role in object perception as well as for motor control. A well-known example is the ability to perceive an object and/or its surroundings as being at rest and clear without blurring when the eyes are moved. This is due to the copied signals relating to self-generated movement being compared with the signals obtained from vision (Holst, 1954). Without this system, we could not perceive an

Interestingly, the forces generated when lifting an object correlate with object weight (Johansson, 1996) and that of heaviness perception (Harper & Stevens, 1948; Stevens, 1958). However, as to the correlation between the forces generated and perceived heaviness, researchers differ. Some researchers support their correlation (H.E. Ross, 1969; Davis & Roberts, 1976; Gordon et al., 1991) and others report their dissociation (Flanagan & Beltzner, 2000; Grandy & Westwood, 2006; Chang et al., 2008). What factors give rise to these

Rinkenauer, et al., 1999) and object torque (Kinoshita et al., 1997).

obtained from the real world (i.e., Real Haptic + Real Vision). Also, we posit that superimposing computer-generated graphical objects precisely onto real, physical objects (i.e., Real Haptic + Virtual Vision) allows manipulation of visual properties of objects independently from physical properties. Finally, we introduce Virtual Haptic + Virtual Vision conditions, using a unique experimental augmented environment using both computer haptics AND computer graphics displays. This allows for selective experimental manipulations in which the haptic device affects haptic perception while computer generated graphics have an effect on visual perception, by superimposing computergenerated object forces precisely onto the graphical objects existing within the computer display. Presented are fresh findings regarding heaviness perception that enhance the knowledge base for the human perception of heaviness. These basic findings are expected to facilitate future research and development of haptic and graphic computer systems relating to human recognition, lifting, transport or manipulation of physical objects.
