2013 13th International Conference on Control, Automation and Systems (ICCAS 2013) Oct. 20-23, 2013 in Kimdaejung Convention Center, Gwangju, Korea

Design of a Haptic Cane for Walking Stability and Rehabilitation

Muhammad Raheel Afzal, Irfan Hussain, Yasir Jan, J ungwon Y oon *

School of Mechanical & Aerospace Engineering and ReCAPT,

Gyeongsang National University, Jinju, 660-701, Republic of Korea

(E-mail: rahee1379@gmail. com.irfanmechatronics@gmail. com.yasirjan@gmail. com.jwyoon@gnu. ac. kr *)

Abstract: Rehabilitation is suggested to be achieved by natural walk, and it may require assistive devices. Assistance provided should motivate the patient to use his own muscle strength rather than be dependent upon the device. Therefore, the devices should only provide minimum support required for the safety, stability, confidence building and guidance. These can be achieved with light touch cue provided at the patient's hands. The proposed haptic cane design has an active haptic handle that can give light touch cue depending upon the body orientation sensed through smartphone. The active haptic handle can be manipulated by a Pantograph mechanism. The Pantograph and arm support's positions and orientation are adjustable. The handle and arm support are mounted on the cane having a single wheel, coupled with motor, shaft encoder and an active brake, for achieving a controlled movement. The proposed design will be able to provide rehabilitation and postural stability for the patients.

Keywords: Rehabilitation, Stability, Haptics, Pantograph, Walking assistive devices.

1. INTRODUCTION

Rehabilitation is a process of improving the capacities of humans who have functional disabilities like full or partial neuromuscular or muscular control of a part of the body. It is important practice to assist a patient through freedom of walk, with the help of walking assistive devices for lower limb muscles. The assistance maybe in a passive form or could have active response towards the body control. Various designs have been proposed for passive walking assistive devices. Hirata, Y. et al. [1] developed a Robot Technology Walker (RT Walker) based on passive robotics that assists the elderly, handicapped people, and the blind that have difficulty in walking. The RT Walker consists of a support frame, two casters, two wheels equipped with servo brakes, and it has passive dynamics that change with respect to applied force/moment. Suzuki, S. [2] developed an intelligent passive IP Cane consisting of mobile base with servo brakes and the cane. The user can move this system based on actual force/moment applied to it. Later on Suzuki, S. et al. [3] developed the concept of the cooperative walking support system, consisting of wearable-type and cane-type walking support systems. The passive designs are mostly consisted of simple base with brakes and as a result the person does not use his body muscles completely, but instead takes full support from the cane/walker. Therefore, the rehabilitation, or activation of weak muscles does not take place effectively.

Passive devices don't usually make a complex decision or does not have any intelligence of its own. Toshio, F. et al. [4-10] developed an intelligent cane robot (iCane) for aiding the elderly having weak lower limb muscles. In the design, a commercial omni-directional wheels robot was used as an omni-directional mobile base, and an aluminum stick was installed on the base of cane robot. The universal

978-89-93215-05-295560/13/$15 @rCROS 1450

joint driven by two DC motors was designed to control the posture of the stick. They proposed a concept called "intentional direction (ITO)" for estimating the user's walking intention. It consists of a 6-axis force/torque sensor fixed to the handle of stick, for analyzing the signals. This approach of walking assistive device uses a complex system for assistance and with a less chance of rehabilitation, because user can put enough body weight on the system which doesn't encourage muscle activation.

Mostly assistive canes provide mechanical support to the old and weak people, but other than that there exists canes which assist the blind people in walking. The blind people need proper feedback from the cane with the help of which they can be guided into some specific direction. The feedbacks are mostly haptic feedbacks giving the blinds a feeling of vibration on hands, in the required direction. Gallo, S. et al. [11] proposed an instrumented handle with multimodal augmented haptic feedback. That handle can be integrated into a conventional white cane to extend the haptic exploration range of visually impaired users. Active feedback based canes also need surrounding information to make some decision. The information is collected through various sensors connected to the canes. Ando, T. et al. [12] also developed a cane with a built-in haptic interface for use by visually impaired persons. The direction and guidance system is based upon the information collected from ICs buried underground. Wang, Y. et al. [13] developed (HALO) system which is a portable and affordable attachment to traditional white canes. The device aims to alert users of low-hanging obstacles by pairing distance data acquired from an ultrasonic range sensor with vibration feedback delivered by an eccentric mass motor. HALO system performs its tasks without interfering with the standard functionality of a white cane. These designs use haptic feedbacks in form of some mechanical vibrations provided to the blind users. These studies are directed towards using haptics for

assistance in navigation; however we propose a concept of using haptics for stability in walking.

Researchers found that using haptic light touch cue for the stability of standing improves the balanced posture. Holden et al. [14] studied how sensory-motor information can be used to stabilize balance in the absence of vision. The sensory motor information about body displacement is provided by contact of the index finger with a stationary bar. Normal body gets stabilized using vision, but in the absence of vision, equivalent stabilization can be achieved at contact force levels less than I N. This amount of force is much less than the actual physical force required for significant physical stabilization of the body. Hence it proves the concept of light touch, as assistance for human brain to produce better upright balance. Kouzaki et al. [15] made a better observation of this concept. They investigated the light touch effect on postural stability during quiet standing with and without somatosensory input from the fingertip, and observed that sway reduction with light touch disappeared when the hand was anaesthetized using a compression block on the upper arm. Experimental study by Boonsinsukh, R. et al. [16, 17] verified that a light touch cue can be provided during walking through the use of a cane. This augmented sensory information provides lateral stability during walking for subjects with stroke by facilitating the activations of weight-bearing muscles on the paretic leg during the stance phase, resulting in better stability when the paretic leg supports the body weight and increased muscle activation for rehabilitation.

In this research, we initially verified through experiments that how stability is achieved with haptic feedback system using Phantom Omni device, as discussed in Section 2. In Section 3 we are proposing the design of a concept for a haptic cane which provides light touch cue during walking, for stability and rehabilitation. Section 4 discusses how the active haptic guidance as a light touch cue is achieved through Pantograph haptic device, similar to the Phantom Omni used in experiments and in the end section 5 is conclusion and future work.

2. EXPERIMENTAL STUDY TO FIND EFFECTS OF ACTIVE HAPTICS

The stability achieved through active haptic feedback system is verified through various experiments. In the experiments we measured body sway, i. e. tilt in ML and AP using smartphone connected to a person's waist. The android application on smartphone, detects angles using internal sensors and sends the information to the computer connected with Phantom Omni device, through wireless network. That computer program continuously reads the data packets coming from the smart phone and gets the ML and AP values of the persons' body at each instant. Fig. I shows the hardware setup used to perform the experiment.

The application running on the computer makes a decision for the direction and force control of the haptic feedback device. The directional forces are generated in

145 1

the x and z axis, but its movement is constrained in y axis. The output force from the phantom Omni haptic device was always limited to be less than IN.

Fig. 1. Hardware setup for experimentation

Various tests were performed on the subjects regarding their stability analysis. The experimental setup was made such that the person is more dependent upon the haptic feedback rather than natural feedback systems. Therefore eye mask was used for limiting visual input. High density foam with dimensions of 600x600x150 millimeters was used to produce soft ground conditions, for certain tests, causing more unstable surface for upright standing.

The tests were of four types, based upon the subject's body postures and ground conditions. PI: Standing using one foot on ground and head

facing forward. P2: Standing heel to toe (tandem Romberg stance) on

ground and head facing forward P3: Standing on one foot on foam and head facing

forward P4: Standing heel to toe (tandem Romberg stance) on

foam and head facing forward Each of these posture tests were taken with three

different feedback categories of haptic response as described below. Fl: No feedback: The subject did not have any

contact with Haptic Interface Point (HIP), i. e. both arms were free and near subject's body.

F2: Directional sensing feedback: The subject's contact to HIP moves in a linear way in the direction of tilt, providing information to the subject's hand about body tilt

F3: Active guidance feedback: The subject's contact to HIP is providing an active guidance feedback force to subject's hand for stability.

So a total of 12 tests (4 postures x 3 feedback types) per subject were taken.

This experiment was performed on a total of 10 healthy participants out of which 7 were male and 3 were female. Participant's age range was between 21 to 32 years. The graphs (in Fig. 2) show the ML (x-axis) and AP (y-axis) values plotted w. r. t. each other, in posture PI and all three feedbacks, of one person.

The dots show the tilt angle of the body and lines show the path followed while moving from one position to another. In each graph, initial position is marked as 0

deg ML and 0 deg AP, and with a set of 300 samples the graph is plotted with the followed path.

No Feedback

f:[�. -4 -3 -2 -; 0 1 2 3 4 5 6

Mediolateral Sway(Degrees)

[l� 1 -8 -7 -6 -5 -4 -3 -2 -; 0 1 2

Mediolateral Sway(Degrees)

Active Guidance Feedback

f ;[ � i0fiG0¥\H . • -2 -1.5 -; -0.5 0 0.5 1 1.5

Mediolateral Sway(Degrees)

Fig. 2. Graph of ML and AP sway in Posture Pl for all three feedbacks.

Other than the graphical analysis, the data was also analyzed numerically to sum up the total body sway (TBS) in each test. Total body sway is calculated by adding the change in tilts after each sample as compared to previous sample. i. e.

TBS = I Ipst - cstl, for all 300 samples, (1)

Where, pst is previous sample tilt angle and cst is current sample tilt angle.

All the experiments led to a result that body sway can be reduced with active guidance feedback from a haptic device compared with providing only haptic sensory feedback.

3. DESIGN OF HAPTIC CANE

The human-haptic interface system is verified through experiments, as discussed previously. Now a practical design is proposed for stability and rehabilitation through a cane with haptic feedback. The complete CAD model, designed in ProlE, of the haptic cane (fig. 3) is explained below.

The haptic cane has the desired mechanical features providing stability and rehabilitation. Stability is achieved by light touch cue, as previous study shows that light touch can assist human brain to produce better upright balance and can also provide lateral stability during walking. Rehabilitation can be achieved because the person does not take full support of cane but uses his own muscles. Mechanical design also consists of an active brake in the single wheel of the cane, with the help of which a controlled movement can be achieved. Similarly a safety mechanism for preventing fall or

1452

serious injury is incorporated into the cane. The proposed haptic cane is design for various

functionalities. The haptic cane is assisted with a smartphone device mounted on the user's waist, providing the sensory feedback for stability.

Fig. 3. CAD model of Haptic Cane

It consists of a single wheel coupled with motor, shaft encoder and an active brake that allows natural walking with various functionalities e. g. variable gait velocity, turning and inclined movements etc. The wheel velocity is measured by using the shaft encoder and turning is achieved by using the gear mechanism with motor on the main shaft of the cane as shown in Fig. 4. In contrast to white canes, it is in contact with the ground all the time, providing more confidence and the ground frictional information to the patient under rehabilitation.

Fig. 4. Wheel and Turning Mechanism

The haptic cane has an active haptic handle which is used for providing guidance feedback. Handle and arm support's position and orientation are adjustable considering the patient's ergonomics. The handle can move with 2-DOF using concept of Pantograph robot, which is a 5-bar mechanism as shown in Fig 5.

The details of Pantograph are explained in section 4.

4. ACTIVE HAPTIC HANDLE

The active haptic handle motion is achieved by using concept of 2-DOF Pantograph robot which is a 5-link mechanism. The parallel design allows larger end-effector forces as compared to a serial manipulator. Linl<- 5 is grounded, joints I and 5 are actuated, and joints 2 and 4 are passive.

The motors and sensors are at point PI and P5 (Fig 6. ). The configuration of the structure can be determined by angle 81 and 85. The force at the handle can be determined by the torque applied by two motors at point PI and P5. The kinematics and jacobian matrices of the Pantograph are previously explained by Campion et al. [ 18] .

Fig. 6. 5-bar Pantograph Mechanism

The angles shown in figure 5, are calculated using the given equations 2-7 of inverse kinematics,

(2)

(3)

1453

(35 =

(4)

(5)

(6)

(7)

The handle is located at point P3 and moves in a plane with 2 DOF with respect to the ground link. In our application, we need the workspace of 8cmx8cm. Considering the workspace required and the ergonomics of the human hand, the link lengths chosen are 4 cm for ground linl<- and 8cm for rest of the linl<.s.

By using the inverse kinematics, it is shown that values of angle a1 and (35 are positive for the whole desired workspace. Fig. 7 shows the desired rectangular workspace of dimensions 8x8 cm, shaded blue.

Workspace of Actiw Haptic Handle

15 - � ------� ----- � ----- � ------ � --

E 10 � UJ

.� >-

o

-10 -5 o X-axis (cm)

5 10

Fig. 7. Workspace of Active Haptic Handle

The linl<- orientations relevant to the handle position at each corner point of the workspace, is shown in Fig. 7. It can also be graphically visualized that the links are in a feasible orientation in the whole desired workspace.

5. CONCLUSION AND FUTURE WORK

The proposed design will be able to provide rehabilitation with the light touch cue during natural walk, and posture stability with haptic handle, based on smart phone sensory feedback. Moreover the design incorporates the safety measures and human ergonomics, which makes it a very practical design. Future work that needs to be done in continuing this research work is the fabrication and implementation of the proposed design.

Acknowledgments

This work was supported by National Research Foundation Korea (NRF) (2011-00313832010-0029690 & 20l2RIA2A2A01047344) and BK21 PLUS(Brain korea21).

REFERENCES

[1] Hirata, Yasuhisa, Asami Hara, and Kazuhiro Kosuge. "Motion control of passive intelligent walker using servo brakes. " Robotics, IEEE

Transactions on 23. 5 (2007): 981-990. [2] Suzuki, Shinji, Yasuhisa Hirata, and Kazuhiro

Kosuge. "Development of Intelligent Passive Cane controlled by servo brakes. " Robot and Human Interactive Communication, 2009. RO-MAN 2009.

The 18th IEEE International Symposium on. IEEE, 2009.

[3] Suzuki, Shinji, et al. "Walking support based on cooperation between wearable-type and cane-type walking support systems. " Advanced Intelligent Mechatronics (AIM), 2011 IEEE/ASME

International Conference on. IEEE, 2011. [4] Huang, J ian, et al. "Study of fall detection using

intelligent cane based on sensor fusion. " Micro-NanoMechatronics and Human Science, 2008. MHS 2008. International

Symposium on. IEEE, 2008. [5] Huang, Jian, et al. "Motion control of

omni-directional type cane robot based on human intention. " Intelligent Robots and Systems, 2008. IROS 2008. IEEE/RSJ International Conference

on. IEEE, 2008. [6] Di, Pei, et al. "A novel fall prevention scheme for

intelligent cane robot by using a motor driven universal joint. " Micro-NanoMechatronics and Human Science (MHS), 2011 International Symposium on. IEEE, 2011.

[7] Di, Pei, et al. "Motion control of intelligent cane robot under normal and abnormal walking condition. " RO-MAN, 2011 IEEE. IEEE, 201l.

[8] Fukuda, Toshio, et al. "Advanced service robotics for human assistance and support. " Advanced Computer Science and Information System

(ICACSIS), 2011 International Conference on. IEEE,2011.

[9] Di, Pei, et al. "Optimal posture control for stability of intelligent cane robot. "RO-MAN, 2012 IEEE. IEEE,2012.

[10] Wakita, Kohei, et al. "Human-Walking-Intention-Based Motion Control of an Omnidirectional-Type Cane Robot. " (2013): 1-12.

[11] Gallo, Simon, et al. "Augmented white cane with multimodal haptic feedback. "Biomedical Robotics and Biomechatronics (BioRob), 2010 3rd IEEE RAS and EMBS International Conference on. IEEE,2010.

[12] Ando, Takeshi, et al. "Development of a cane with a haptic interface using IC tags for the visually impaired. " Intelligent Robots and Systems, 2009.

IROS 2009. IEEE/RSJ International Conference on. IEEE, 2009.

[13] Wang, Yunqing, and Katherine J. Kuchenbecker. "HALO: Haptic Alerts for Low-hanging Obstacles in white cane navigation. " Haptics Symposium (HAPTICS), 2012 IEEE. IEEE, 2012.

[14]

[15]

[16]

[17]

[18]

1454

Holden, Maureen, Joel Ventura, and James R. Lackner. "Stabilization of posture by precision contact of the index finger. " J Vestib Res 4. 4 (1994): 285-301. Kouzaki, Motoki, and Kei Masani. "Reduced postural sway during quiet standing by light touch is due to fmger tactile feedback but not mechanical support. " Experimental Brain Research 188. 1 (2008): 153-158. Boonsinsukh, Rumpa, Lawan Panichareon, and Pansiri Phansuwan-Pujito. "Light touch cue through a cane improves pelvic stability during walking in stroke. " Archives of physical medicine and rehabilitation 90. 6 (2009): 919-926. Boonsinsukh, Rumpa, et al. "Clinical identification for the use of light touch cues with a cane in gait rehabilitation poststroke. " Topics in Stroke Rehabilitation 18. 1 (2011): 633-642. Campion, G. ; Qi Wang; Hayward, v., "The Pantograph Mk-II: a haptic instrument," Intelligent Robots and Systems, 2005. (IROS 2005). 2005 IEEEIRSJ International Conference on , vol. , no. , pp. 193, 198,2-6 Aug. 2005 doi: 10. 11 09/IROS. 2005. 1545066