Search

Purpose: To explore the challenges and strengths of employing an extension approach, which entwines indigenous knowledge and science, for the purpose of enhancing sustainable agricultural practices. Design/methodology/approach: We assessed an agricultural extension approach which entwined indigenous knowledge and science. Data for this paper were qualitative and quantitative in nature as gathered from semi-structured interviews, farmer observations, surveys and the researcher's field notes. Findings: Indigenous knowledge has enormous potential to contribute to sustainable agriculture, yet it is rarely included in agricultural extension. There are many inhibiting factors which may contribute to the lack of inclusion of indigenous knowledge in extension programmes, including: perceived value of indigenous knowledge; knowledge protocols; cultural constraints; access to indigenous knowledge; and, misrepresentation of indigenous knowledge, but also strengths such as enhanced sustainable agricultural practices. Practical Implications: Understanding the strengths and challenges of entwining indigenous knowledge and science may support extension policies which further value indigenous knowledge and recognise accessibility and property rights, thus maximising the strengths and limiting the challenges. Theoretical implications: Whilst the present study further supports theories which recognise the challenges (accessibility; perceived value; intellectual property rights; cultural embeddedness) and strengths (contribution to sustainable agriculture) of including indigenous knowledge in agricultural extension, this study also offers an alternative approach to agricultural extension. Originality: Much has been written regarding the value of indigenous knowledge and its potential in sustainable agriculture. This paper explicitly reports on the strengths and challenges of entwining indigenous knowledge and science in an alternative agricultural extension approach.

Engineering design is evolving into a global strategy that distributes engineering effort to team members around the world. Because modern engineering design uses analytical models, model information must be distributed globally through computer networks. This strategy would be improved if component suppliers were able to efficiently provide dynamic models of supplied components. Furthermore, to use these component models, they must be efficiently assembled to obtain a dynamic model of a product using them. Four characteristics are needed to enable this distribution and assembly process. These characteristics are a unique standard model format, an exchange of model information through a single-query network transmission, external component models protecting proprietary internal design details, and, finally, a recursive assembly process. The modular model assembly method (MMAM) (Radcliffe et al., 2009, 'Networked Assembly of Mechatronic Linear Physical System Models,' ASME J. Dyn. Syst., Meas., Control 131, p. 021003) is a model assembly algorithm that satisfies these requirements. The MMAM algorithm assembles linear physical system models with dynamic stiffness matrices. In an affine system, deviations in the inputs and outputs exhibit a proportional relationship, but the outputs of the system are nonzero at zero input (Buck and Willcox, 1971, Calculus of Several Variables, Houghton Mifflin, Boston). One motivation for developing a process to assemble affine systems is the wide use of such models resulting from local linearization of general differentiable nonlinear physical system models about a nonzero, but constant, operating point. This paper provides the first general approach to the 'operating point problem,' where the operating points of each individual component are solved as a function of the desired operating point of the model of an assembly of those components. The solution of this problem allows the assembly of linearized system models at any requested system operating point. This paper extends the MMAM to nonlinear affine system models. The MMAM uses internet agents to provide external models of components when requested by either users or other model agents. Assembly agents use the models provided by component agents to build an analytical model using models provided by component agents and assembly constraints within the assembly model agent. MMAM models are supplied in a standard form that allows an assembly agent to put together efficiently a model of the assembly that is also in the standard form. The process is recursive and facilitates hierarchical use of agents to efficiently build assemblies of assemblies to any level of complexity. [DOI: 10.1115/1.4002471]