About

The breadth of skillsets that are required to properly train a new cadre of workforce makes curriculum design and integration with existing curricular frameworks in traditional ECE departments incredibly challenging. For a holistic appreciation of the entire neuromorphic engineering/computing paradigm, students need to be adequately trained in introductory concepts of neuroscience (neuronal models and dynamics), electrical modeling, material science, device physics/characterization, and learning algorithms.

To address this issue, we propose the NeuRoBots educational consortium among the partnering institutions. The main objective of this consortium is to collaboratively develop and implement a comprehensive workforce development plan that incorporates evidence-based best practices to help train a new generation of practitioners and researchers. 

Our proposed research addresses challenges at all levels, from choice of materials and devices with desired properties, circuits and novel architectures, and algorithms to enable robots that can learn to move. The novelty and challenges in our proposal require a holistic academic background encompassing diverse areas such as material science, device physics, circuit modeling and simulation methods, hardware prototyping, software development, control theory, and learning algorithms. However, a truly integrated instructional platform bridging these areas is currently lacking. The cornerstone of our plan is a self-contained instructional platform, comprising lecture videos, PowerPoint slides, and laboratory manuals, which will expose students to the above topics. Examples of specific components of our planned curriculum include: (i) (UW) materials modeling and atomic simulation of novel materials, (ii) (UPitt) device modeling, fabrication, and characterization, (iii) (UPitt) IC design, simulation, testing and verification, and (iv) (UPitt) embedded system design and validation for robotics.

Research outcomes from this proposal will be integrated into the curriculum for undergraduate and graduate courses taught by the PIs, design projects, and summer internships for high school students. Each instructional module will have clearly identified learning objectives, mapped to our application domain (robot locomotion), to clearly illustrate the applications of conceptual/theoretical tools. Further, each module will be followed by a set of assessment questions requiring creative thought. Practical issues will be emphasized so that students are cognizant of the challenges they will encounter in their professional lives. Our planned instructional media, which will be carefully modularized to facilitate its adoption in formal undergraduate courses as well as specialized certificate courses on hardware development and neuromorphic engineering, will enable students to gain a holistic view of the product development stack.

Curriculum development activity will be led, developed, assessed, and refined in coordination with our proposal partner at EWU. Das at EWU will hire two undergraduate students who will assist him with preparation of the instructional material and organizing outreach and dissemination activities geared toward students at high schools, and community & 4-year colleges. He will also offer a directed study course on Neuromorphic Computing for senior undergraduates in Computer Science and Electrical Engineering students at EWU. Undergraduate students will be encouraged to pursue senior capstone design projects in select areas of the project.

All modules of our educational platform will be critically assessed which will provide feedback on student progress and the need for refinement of those modules. Our primary assessment mechanism will be anonymous pre- and post-surveys wherein students will provide feedback on the individual modules.

The ultimate outcome from the work of the NeuRoBots educational consortium will be job-ready students that are equipped to satisfy the growing needs of the semiconductor industry. To assess this outcome, several key, measurable objectives have been identified. First, we will measure the self-efficacy of the students with respect to holistically applying their knowledge to meet the needs of the semiconductor industry. We will work in tandem with our industry partners to develop and administer this survey. This survey will be carried out annually. We will also evaluate the employment preferences and student post-graduation plans via student interviews and focus groups, conducted each year. The combination of these student-centered measures will allow us to evaluate the extent to which the program not only builds competency, but also enthusiasm for the industry. Second, we will gather basic statistics on job applications, placements, for both co-curricular activities and full-time employment. We will also involve key industry stakeholders. We will assemble a group of industry stakeholders to assess the quality of student participants and their suitableness for employment in the field.

Employers will be asked to evaluate student resumes and projects and asked to provide feedback via carefully crafted rubrics. This will allow the investigators to have data on job readiness, irrespective of job placement, which can, in some instances, have a time horizon that is longer than the duration of the research program. To ensure that the aforementioned items are specifically tailored to the semiconductor industry, the workforce evaluation will be carried out by a faculty member at the University of Pittsburgh who has expertise in both semiconductors and program assessment. Profs. Samuel Dickerson and Renee Clark at the Engineering Education Research Center (UPitt) have both agreed to assist us with crafting an assessment plan.