Vocational Education Technology: Rural India Bhavani B. Amrita Vishwa Vidyapeetham School of Engineering Amritapuri, India bhavani@amritapuri.edu Srividya Sheshadri Amrita Vishwa Vidyapeetham School of Engineering Amritapuri, India srividyasheshadri@am.amrita.edu Unnikrishnan R. Amrita Vishwa Vidyapeetham School of Engineering Amritapuri, India unnikrishnanr@amritapuri.amrita.edu ABSTRACT Vocational Education and Training (VET) helps bridge the gap between limited education and gainful employment. However, it is difficult to make VET readily accessible to economically and educationally challenged communities. To meet this challenge, we introduce a novel solution that adds unique elements to the realms of education technology: vocational education enhanced through both multimedia and haptic technology. This paper will present conceptual elements of revolutionary haptic technology to be applied in vocational education settings as well as present two different vocational training tested in the field using multimedia and haptic technologies. Through the application of multimedia, virtual reality and haptic technology to VET, we demonstrate that this enhanced vocational education and training has the capacity to convey complex concepts and skill sets to illiterate and semi-literate individuals. We also show the vast potential for applicability of such enhanced VET across India, and the resulting promise of increased employment and improved livelihoods for hundreds of thousands of people. Categories and Subject Descriptors J.0 Computer Applications education technology General Terms Education Technology, Experimentation, Human Factors Keywords Vocational Education Technology; Multimedia Applications; Haptic Technology; Computerized Vocational Education and Training 1. INTRODUCTION Vocational skill development is linked to both economic productivity and social wellbeing [International Labor Organization 2008]. The implications of this inherent principle are vast in a country such as India, where nearly a third of the one billion plus nation lives below the poverty line. Vocational education and training (VET) provides an avenue by which individuals - particularly those on the sidelines of economic growth - can advance both economically and socially. "Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. A2CWiC 2010, September 16-17, 2010, India Copyright 2010 978-1-4503-0194-7/10/0009 $10.00" However, vocational education schools have been consistently unsuccessful in penetrating the communities that would most benefit: the economically marginalized, and particularly those residing in rural areas where unemployment and limited education are most prevalent [Majumdar 2008]. Communities with limited education and opportunities for skill development become the perfect environment for the introduction of accessible vocational education technology. Flexible, accessible VET technology has demonstrated effectiveness in transcending economic, geographic, and academic barriers so common to traditional education. This paper will discuss how the project Sakshat Amrita Vocational Education (SAVE) brings vocational education to rural India utilizing multimodal computer interfaces that interact with the user using visual, haptic and auditory feedback. Furthermore, we will demonstrate how this novel combination of Multimedia with Virtual Reality & Haptic technologies addresses the limitations of traditional education; consequently enhancing the employment prospects of India s economically disadvantaged populations. The discussion will begin with an exploration of existing education technology specifically utilizing multimedia and haptic technologies. SAVE is subsequently introduced as a novelty in the field of education technology and vocational education specifically, given its incorporation of both multimedia and haptic technologies. In this paper we will present haptic applications currently at the final stages of development and two completed modules of vocational training that were tested in the field. While SAVE envisages the development and deployment of various computerized vocational education training courses over time, the first implementation was Fabric Painting and the introductory module of Household Plumbing. We will illustrate the technical architecture of the SAVE framework and conclude the discussion with an impact analysis of SAVE s initial test cases and planned future testing of haptic technology in development. 2. EDUCATION TECHNOLOGY 2.1 Multimedia In a randomized control experiment of 74 occupational physicians, e-learning was determined as equally effective in knowledge delivery as traditional lecture instruction [Hugenholtz, N. et al, 2008]. The results of this study carry
significant implications for resource-sparse areas, where knowledge can just as effectively be transferred to a large number of people through minimal resources and infrastructure if appropriate e-learning tools are used. Kher Hui Ng and Ryoichi Komiya [2002] identified the advantages interactive multimedia textbooks have over standard textbooks. 15 undergraduate students and university staff were selected at random to test and assess the efficacy of multimedia textbooks. The multimedia textbooks were found to be advantageous in reducing average learning time, raising interest in the subject matter, and in enhancing understanding of subject matter. In addition, the multimedia textbook better accommodated varied learning styles. The latter is particularly informative when considering a vocational education scenario where the range of learning aptitudes and styles is the norm. A multimodal representation of material via multimedia education technology has also demonstrated various benefits [Price, 2007]: the promotion of active learning compared with traditional methods that often engage students in passive learning; the reiteration of material in various ways made possible through the multi-media application, which offers multiple avenues by which the student can grasps educational content; accessibility, as the technology is geared toward varied learning styles as well as learning abilities (ie. Auditory and visual learners as well as the visual and hearing impaired). 2.2 Haptic Education Technology Haptic technology immerses the user in a virtual world through the sense of touch, and has become the subject of great interest. Most cited cases thus far are in the medical field particularly in the areas of surgery, laparoscopy, endoscopy and dental simulators [Coles, T., Meglan, D., John, N. 2010; Körner, O., Männer, R.2003]. Apart from the application of haptics in medical training, other notable research efforts applying haptics in educational technology include Hamza-Lup & Adams HaptEK16 module (Haptic Environments for K16) [2008] for teaching Pascal s principle in Hydraulics. Students who learned Hydraulics principles using the haptic hydraulics simulator were found to have a better understanding of the concept than other groups of students that were not exposed to the haptics simulator. Salisbury et al's [2007] seminal work on evaluating the role played by haptics in force skill training further underscores the need for the use of haptics in vocational training. These authors found that among three groups of volunteers, each assigned to a haptic, visual and a visuo-haptic exercise, the group that learned a sequence of forces in one dimension space using visual and haptic modalities had better force recall. Since most vocational trades require skills that emphasize a sequence of force actions, a visuo-haptic paradigm would be best suited to the task. While the aforementioned case studies demonstrate the significant impact technology makes through education, little research has been done to date on the particular impact educational technology has on the delivery of vocational education and training. This is largely due to the fact that few models have attempted to extend the reach of education technology to vocational education and training, particularly within India s academically deprived and economically impoverished majority. 3. S.A.V.E As demand in labor continues to grow, traditional vocational education and training (VET) schools face challenges of limited trainers, time and materials per pupil [Guo, G. Zhan, Q., Liu, G. 2009]. Consequently, the quality of instruction tends to degenerate and the prospects of actually becoming employed are also diminished. In rural India for instance, vocational education reaches only 3 percent of employable youth, and there are few signs they have even seen any benefit from this training [ILO, 2006]. SAVE accordingly offers a solution that addresses all of these limitations. Given the demonstrated effectiveness of multimedia and haptic technologies in related fields of education technology, SAVE s approach to automating vocational education via Virtual Reality, Multimedia & Visuo-Haptic technologies anticipated the following results and possible applications: 1) Decreased dependency of expert trainers in vocational training, and greater reliance on standardized computer-based evaluation methods. 2) Reduction in time to complete training as courses are self-paced and user driven. 3) Greater exposure to a practically infinite number of technology-driven learning scenarios, which will in turn diminish consumption of more traditional training materials (ie. raw materials such as wood in the case of wood carving). 4) Standardization of vocational training across India. 3.1 SAVE Technologies in Development The following haptic technologies are in the final stages of development for the plumbing module, which will specifically address the requirements of vocational education and training. 3.1.1 Visuo-Haptic Workshop Environment CHAI3D, an open source visuo-haptic engine, was used to develop the workshops. CHAI3D is a 3D scene-graph library with real world deformable & rigid body physics engines integrated into it. Graphics in CHAI3D are handled by OpenGL, which offers flexibility in adding external particle engines and OpenGL code to customize existing features. CHAI3D allows for 3D objects built using commercial design software such as Maya and 3dMAx to be loaded inside the virtual environment. This enhances the immersiveness factor as the users find that the virtual objects very closely approximate the look and feel of real world tools and workshop materials.
animation of the cut piece falling down is displayed signaling a successful exercise. Figure 5. Basic architecture for a virtual reality application incorporating visual, haptic & auditory feedback [Salisbury et al, 2004] 3.1.2 Novint Falcon Originally engineered as a gaming device, the Novint falcon offers three degrees of freedom (DOF) as well as three degrees of force feedback (http://home.novint.com/products/novint_falcon.php. It is currently the cheapest multi-dof haptic device available and is the primary means for developing our initial prototypes. It has a workspace of dimension 4" x 4" x 4" and gives a maximum force feedback of 9 N. However, one major disadvantage of the Falcon is that the workspace area is limited and feedback is too low for simulating vocational tools. This led to development of low cost alternative haptic devices. 3.1.3.5 Custom Falcon Grip To increase the immersiveness of the application, we replaced the standard grip of the Falcon with our own custom made grip that came close to that of a normal hacksaw (Figures 7a and 7b). Figure 7a. The custom grip interacting with the pipe cutting application 3.1.3 Low Cost Haptic Devices After determining that the accessibility of these aforementioned haptic workshops is directly correlated to the cost and availability of the haptic devices, we started developing cheap haptic devices with limited degrees of freedom. These devices are designed to be even cheaper than the Falcon, but will provide more realistic force feedback than the Falcon. The design process of the device took into account factors such as workspace dimensions, the number of degrees of freedom, force experienced, etc. The force experienced by the student in real life is measured using load sensors, and these measurements went into the design process to create hardware that could recreate these forces. The first haptic device under development has two degrees of linear motion and three degrees of rotation. This will be used to simulate tools like the hacksaw, file, and more importantly, more expensive tools like the band saw and jigsaw. 3.1.3.2 Pipe Cutting Pipe cutting forms an important part of any plumbing activity. Here the emphasis is on a linear bi-directional movement with the hacksaw on the surface of the pipe. The critical factor to be considered by the user is that the hacksaw should be straight and perpendicular to the pipe surface and never at an angle. For this we devised a novel grip mechanism for the falcon (Figures 7a and 7b). 3.1.3.4 Mesh Cutting Collision detection algorithms in CHAI3D detect the mesh triangles in contact with the hacksaw and mesh cutting is applied to split the pipe into two as the cutting operation proceeds. The realistic haptic force feedback for cutting is emulated using the stick and slip force effect available in Chai3D. Once cut, the meshes are separated and an Figure 7b. The Novint Falcon with the custom grip for pipe cutting The custom grip has an embedded accelerometer that communicates with the software through a serial connection. Here we use a triple axis accelerometer (MMA7260Q). The serial communication thread takes this data from the serial port of the computer and converts the signal into corresponding angular values using a look up table. These values are then applied to the rotation matrix of the hacksaw in the graphics thread so that as the custom grip rotates, the hacksaw turns. The application has two modes of operation. One deals with training the student with the operations involved in cutting a pipe, and the other evaluates the student s performance. Training mode: Here the user is given step-wise instructions for cutting a pipe and is allowed to make mistakes and adjust responses accordingly. Evaluation mode: Feedback in the form of error messages are generated, warning the user if the tilt angle of the hacksaw goes beyond 10 degrees. The application evaluates the student s performance and assigns a score based on the time taken, mistakes made and their severity. These results
are then sent to the main SAVE application to keep track of the user s progress. 3.1.3.6 Filing Filing is a material removal process used in carpentry and metalwork normally for finishing operations to smoothen any given surface. The purpose of this workshop is to create the look and feel of a filing operation with visuohaptic as well as auditory feedback (Figure 8). (calculated using the God Object algorithm), d t is calculated and the pitch of the recorded audio sample is varied accordingly. The proportionality constants for these relations (Fig 10 (c)) are arbitrarily assigned for now pending further study of metal properties. Figure 8: The Filing Application Graphics The virtual metal s surface texture is changed dynamically upon contact with the filing tool. This makes use of OpenGL s Frame Buffer Object (FBO), which stores all pixel data for a texture (Figure 9a). Once the contact area is known, the corresponding pixels in the texture buffer are modified. A light hue is added progressively to the pixels in contact (Figure 9b). This simulates the visual feel of the actual filing process. Figure 10. Audio Feedback It is anticipated that the aforementioned technologies will serve as prototypes for simulated practical development for other vocations as well such as in carpentry, welding, and wood carving. What follows is a discussion of two vocational training courses: fabric painting and introductory module of household plumbing that have been tested in the field. 4. S.A.V.E ARCHITECTURE The objective to extend vocational courses to areas plagued by unemployment and illiteracy requires a framework capable of enormous scalability. Broadly, the SAVE engine (as illustrated in Figure 11) contains the entire application, developed using Adobe Flex 3.0. Admin Module Figures 9a and 9b. Metallic texture before and after filing Haptic feedback CHAI3D's inbuilt collision detection and force rendering algorithms are used to create force feedback. In this case we define an Axis Aligned Bounding Box (AABB) around the 3D mesh surface of the virtual metal for collision detection since it is the fastest method for detecting collisions in the case of regularly shaped objects like the virtual metal piece. The force feedback is calculated by CHAI3D itself using the God object haptic rendering algorithm [Zilles, C. B. and Salisbury, J. K. 1995]. Audio Feedback To accurately recreate the audio effect of filing, the area of the filing tool in contact with virtual metal and the force exerted downwards are measured for each discrete time period t (Fig. 10). The difference between the areas (a t+1 and a t ) for two consecutive time periods and the distance between the proxy point and Haptic Interface Point General Design Components Tutorials Course Course Specific Components Help Chapters Classroom Interactive Workbook Haptics Workshops Evaluation Figure 11. Macro view of the Multimedia Vocational Education Platform The Engine contains two major functionalities: the Administration Module and the Course Module. The Administration module is designed such that every component in the framework can be accessed and edited from a single interface. The administrative user can update any content visible on the application screen for example adding, editing or removing courses, course content, or more general components such as background imaging, tutorial videos, etc. The aim of reaching all areas of India requires multi-language capability. The Engine accordingly
allows for content revision in every major Indian language, in addition to English. The application is structured modularly to enable flexibility and accessibility to the student. As seen in Figure 12, the multimedia vocational course structure is comprised of four components contained within chapters of the course: classrooms, interactive workbooks, virtual workshops that simulate practical hands on learning through haptic devices, and a tests component where users can self-evaluate their own progress. 4.2 Workbooks Workbooks utilize 2D illustrations and 3D animation to create interactive exercises where users navigate through practical real-world scenarios and apply the theoretical knowledge gained through the Classroom. Figures 14a and 14b illustrate the interactive workbooks of the fabric painting and plumbing courses, respectively. Virtual Classroom and Video Lectures Student Evaluation and Feedback Interactive Workbooks and exercise of theoretical concepts Haptics and Virtual Reality Skill Development Figure 14a. Fabric Painting Workbook Figure 12. Multimedia Enhanced Course Structure 4.1 Classrooms Classrooms (Figures 13a and 13b) provide theoretical knowledge on a specific component in the course through video lectures accompanied by text, audio, 2D and 3D illustrations and animations. Figure 14b. Plumbing Workbook Figure 13a. Fabric Painting Classroom Figure 13b. Plumbing Classroom 4.3 Workshops Workshops house the haptic and virtual learning environment component of the application. 5. METHODOLOGY The fabric painting and plumbing courses were tested in the field to determine the impact on learning. Given the aim of developing vocational education training that transcends current VET system limitations, the target sample population consisted of users concentrated within rural communities of Kerala, India. The target sample population consisted of individuals who have faced (or will face) limited employment opportunities as a result of poor educational outcomes or limited economic resources. Based on these criteria, two sites were selected within the predominantly rural district of Idukki within the state of Kerala, India: Nedugandam, and the Marayoor Tribal Settlement, Kumbitankuzhy. The courses were delivered over a period of 10 days, during which participants received two hours of training through the computerized application followed by two hours of practical hands-on training through a resource expert.
6. RESULTS 6.1 Training One: Fabric Painting While nearly all of the participants had never used a computer, post course evaluations revealed a unanimous feeling of comfort with the technology and application. 70 percent reported they felt the application and digital tool enhanced their learning of fabric painting. Figure 11. Deployment Participants in the fabric painting workshop 6.2 Training Two: Fabric Painting and Plumbing Fabric painting (revised since the first deployment) and Household Plumbing were tested over 10 days. Awareness for the course was generated through a community hall meeting. Given that all of the community members met target population criteria, registration was conducted on a first come, first served basis. Each course consisted of 30 students, 60 participants in total. Participant pre- and post-course surveys were conducted to assess the socio-economic make-up and pre-existing technological capabilities of the sample population, as well as experiences with the courses and other outcome factors (below). 6.3 Pre-Course Findings When community members were presented with the opportunity to take the fabric painting and plumbing courses, 43 percent of participants revealed the reason they took the course was to gain exposure to computers. This response confirmed that computers are an effective medium in generating interest vocational training, no matter the trade. 6.3.1 Educational and Employment Prospects Approximately 32 percent of the participants reported their highest level of education to be between the 6 th and 9 th grades. Of this cohort, over 76 percent reported dropping out of school before completion. Over 50 percent of those employed at the time (49 percent of the participants), revealed they earned an average daily income of Rs. 100 through daily wage labor. 6.4 Post Course Findings While educational attainment levels varied significantly as identified through pre-course analysis, there was minimal variance in participants ability to navigate through the application and course material. These findings confirmed that educational attainment had little to do with ability to navigate through the application, and to grasp new concepts through multi-media presentation. 6.4.1 Perceptions of the Application In order to identify whether specific aspects enhanced the learning experience, participants were asked to select which elements of the application inspired interest within the courses (fabric painting and plumbing respectively). Findings are illustrated in the graph below (Chart 1). Course Content refers to the curriculum content; Multimedia Course Structure refers to the applications as a whole (the compilation of 2D and 3D graphic imagery, animation video lecture and interactive workbook and workshop). The remaining three categories correspond to their respective sections within the course: Classroom, Workbook and Workshop. As demonstrated in the graph below (Chart 1), the multimedia course structure and workshop elements were identified by the participants as enhancements to the learning experience. 15% 20% 25% 17% Chart 1. Participants Perceptions Of particular interest to the developers of the application was whether it stimulated interest in learning, made learning easy, whether users found the video lectures in combination with text helpful, felt the application enhanced their learning, and whether the applications complemented the learning process (as opposed to becoming an impediment). Participants were asked to assess their feedback using a 4-point Likert scale. The results are displayed in Chart 2. 18% 23% 20% 23% 20% Chart 2. Impact on Learning 19% Course Content Multimedia Course Structure Classroom Workbooks Workshops Stimulated interest Learning made easy Text/video helpful Enhanced learning Complements Learning Process
6.4.3 Reduced Training Time Both fabric painting and plumbing took 40 hours to complete (4 hours for 10 days). As compared to the official Government of India curriculum for fabric painting and plumbing, SAVE courses saved 90 hours of training a 73 percent reduction in the time required to train individuals in identical curriculum. In a focus group interview with participants, students revealed a comfort level with the technology and course such that an expert trainer was only required much later in the curriculum. REFERENCES [1] Coles, T., Meglan, D., John, N. 2010. The role of haptics in medical training simulators: A survey of the state-of-the-art. IEEE Transactions on Haptics, IEEE computer Society Digital Library. IEEE Computer Society. [2] Guo, G. Zhan, Q., Liu, G. 2009. Construction of a training model of vocational skills in information technology environment. Computer Network and Multimedia Technology, International Symposium 1-4. 5. FUTURE WORK While the courses received an overwhelming satisfaction rate of over 90 percent (fabric painting and plumbing combined), participants requested greater exposure to the workshop element. As we move into the next phase of development, further integration of haptic technology will arguably strengthen course delivery and address participant feedback. A Learning Management System (LMS) is also in development to analyze user strengths and weaknesses in order to effectively reconfigure the teaching pattern to maximize learning. Such a modification will help developers more precisely evaluate the impact of the multimedia-haptic technology enhanced vocational education platform. We also plan to test the four technologies described conceptually within the following three months. Ultimately 28 vocational courses utilizing multimedia and haptic technology will be developed and deployed. 6. CONCLUSION These initial results demonstrate that vocational education enhanced by multimedia and haptic technology is effective in conveying complex concepts and skills to predominantly semi-literate individuals. These initial results further confirm effectiveness in the following areas: 1) Decreased dependency on expert trainers 2) Reduction in time to complete training 3) Increased exposure to learning scenarios 4) Standardized course content The quantifiable results clearly evidence a positive reception of the application among the target sample population. In addition, it is apparent that the application of multimedia and haptic education technology to vocational education served as a social equalizer, in which people of both genders with varied educational backgrounds were able to learn side by side. Such findings are particularly significant given the mission to deliver vocational education to a massive impoverished and heterogeneous population. ACKNOWLEDGEMENTS The research and development of SAVE is funded by the Government of India, MHRD. We are grateful to the members of the SAVE Lab whose hard work and dedication has made this work possible. 1. [3] Hamza-Lup, F. G. and Adams, M. 2008. Feel the Pressure: E-learning Systems with Haptic Feedback. In Proceedings of the 2008 Symposium on Haptic interfaces For Virtual Environment and Teleoperator Systems (March 13-14, 2008). HAPTICS. IEEE Computer Society, Washington, DC, 445-450. DOI= http://dx.doi.org/10.1109/haptics.2008.4479991 2. [4] International Labor Organization. 2006. Employment challenges and strategies in India. 3. [5] Kikuchi, T.; Kenjo, T. 1996.; Developing multimedia training materials for use with small robot controls at Chubu Polytechnic Center in Japan. Education, IEEE Transactions 39 (3) 349-356. 4. [6] Körner, O., Männer, R.2003. Implementation of a haptic interface for a virtual reality simulator for flexible endoscopy. 11th Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems (HAPTICS'03). 5. [7] Majumdar, S. 2008. Workforce development in India policies and practices. Asian Development Bank Institute. 6. [8] Ng, K.H., Komiya, R. 2002. Multimedia textbook for virtual education environment. Engineering Science and Education Journal 11(2), 73-79. 7. [9] Price, P. 2007. Multimedia technologies and solutions for educational applications: Opportunities, trends and challenges, Multimedia Signal Processing, 2007. IEEE 9th Workshop. 8. [10] Salisbury, K., Conti, F., and Barbagli, F. 2004. Haptic rendering: Introductory Concepts. IEEE Computer Graphics and Applications. 24(2) 24-32. 9. [11] Morris, D., Tan, H., Barbagli, F., Chang, T., and Salisbury, K. 2007. Haptic Feedback Enhances Force Skill Learning. In Proceedings of the Second Joint Eurohaptics Conference and Symposium on Haptic interfaces For Virtual Environment and Teleoperator Systems (March 22-24, 2007). WHC. IEEE Computer Society, Washington, DC, 21-26. DOI= http://dx.doi.org/10.1109/whc.2007.65 10. [12] Zilles, C. B. and Salisbury, J. K. 1995. A constraintbased god-object method for haptic display. In Proceedings of the international Conference on intelligent Robots and Systems-Volume 3 - Volume 3 (August 05-09, 1995). IROS. IEEE Computer Society, Washington, DC, 3146.