Integrating Innovations

​An innovation can be technically feasible and still fail to work in practice. Its viability depends on how it fits with the people, organizations, workflows, incentives, and institutions around it.​​ This stream of work focuses on the next part of the innovation journey: how promising technologies are evaluated, adapted, supported, and integrated into the larger systems in which they must operate.​ This work is especially important in settings such as healthcare and generative AI deployments, where different stakeholders may value different outcomes, where implementation depends on multiple organizations, and where technical performance is only one part of whether an innovation can succeed.

​

Papers

1. Barriers to Inter-organizational Knowledge Transfer in Post-Hospital Care Transitions

Shi-Ying Lim, Sirkka Jarvenpaa and Holly Lanham,  Journal of Management Information Systems, 2015 [Link]​

Post-hospital care transitions require information and knowledge to move across hospitals, patients, and community providers, often under significant time pressure.

This study identifies barriers to inter-organizational knowledge transfer, including difficulties accommodating changing absorptive capacity and reconciling conflicting knowledge and goals. It develops directions for how information systems may help address these barriers in high-risk interorganizational settings.

​

2. Feasibility, Reliability and Validity of Self-Measurement of Knee Range-of-Motion Using a Smartphone Application, Chew, E.S., Woon, E.L., Low, J.K., Haseler, L.J., Ismail, I.*, Alif, M.B.*, Kwan, Y.H., Tan, J.W.M., Koh, S.S.M., Pang, H.N., Chen, J.D.Y., Yeo, S.J., Lim, S.Y., Tan, C.L.C. & Pua, Y.H. , PLOS One, 2024 [Link]

This interdisciplinary project examines whether patients recovering from total knee arthroplasty can independently use a smartphone application to measure their own knee range of motion. We find that patient self-measurement is feasible, reliable, and accurate, demonstrating the potential for patient-facing smartphone measurement to support more accessible rehabilitation and remote monitoring.

​

Other work in progress 

3. Selecting AI Solutions for Population Health: Tensions and Mitigations. Idauli Tamarin* & Shi-Ying Lim, Working paper

The rapid proliferation of health AI creates a difficult selection problem: a technically promising solution may still create tensions with the context in which it will be used or with the different goals stakeholders are trying to achieve. Using a case study of a digital innovation selection process for population health, we identify tensions arising from AI's technical characteristics, its contextual environment, interactions between the two, and competing population-health goals. We also examine how stakeholders attempt to mitigate these tensions during selection and how some responses create consequences that appear only in later stages.

​​

4.  Do “Free” Customers Help? Vendor and User Outcomes in Grant-Mediated GenAI Product Development. Arthur Hu* & Shi-Ying LIM, Working paper

Innovation grants can create relationships between technology vendors and users that might not otherwise emerge through normal market exchange. This project examines what happens when GenAI vendors receive subsidized access to users for product development. Rather than assuming that access to customers necessarily improves learning, we examine how the structure of these grant-mediated relationships shapes outcomes for both vendors and users, including what vendors learn, how products develop, and what participating users ultimately receive.

​

5. Deviance by Design: How Grant Ecosystems Create the Conditions for Their Own Exploitation. Arthur Hu* & Shi-Ying LIM, Working paper

Innovation-support programs are designed to lower barriers, encourage experimentation, and help promising ideas develop. But the same rules and incentives that make participation easier can also create opportunities for strategic behavior that moves away from the program's intended purpose. This project examines how such behavior can arise from the design of the grant ecosystem itself. The focus is therefore not only on participants who exploit a program, but on how the structure of an innovation-support system shapes the behaviors it makes possible.

​

6. Ontological Disagreement in Generative AI: How Technological Attributes Drive Contestation in Anime. Hafizh Rafizal Adnan* & Shi-Ying LIM, Working paper​

This study examines contestation surrounding early GenAI anime projects. We identify three technological attributes—pattern extraction, automated execution, and stochastic generation—and show how community members interpret them through craft-centered, evolution-centered, and utility-centered frames. These incompatible frames produce disagreement over what counts as legitimate learning, effort, and authorship. The study helps explain why transparency, demonstrations, or dialogue may fail to resolve resistance when stakeholders disagree about the basic criteria by which a technology should be judged.​