CXOToday has engaged in an exclusive interview with Prof. Somya Singhvi, Marshall School of Business, University of Southern California, Assistant Professor of Data Science and Operations
Q1: How do you define sustainable operations and socially responsible operations, and why are these concepts particularly important for developing countries?
A: Sustainable and socially responsible operations refer to managing resources, processes, and practices to ensure the long-term viability and health of the business, the environment, and society. This involves minimizing negative environmental impacts while maximizing the efficient use of resources and ensuring economic and social benefits. A particular focus is on conducting business in a manner that is ethical and considers the well-being of all stakeholders, including employees, communities, consumers, and the environment. These concepts are crucial in developing countries because they support sustainable economic growth while addressing pressing social and environmental challenges. By adopting sustainable and socially responsible practices, businesses can contribute to national development goals, enhance their competitiveness, and ensure the long-term viability of their operations by reducing risks and building trust with producers, consumers, and partners.
Q2: Can you elaborate on the challenges faced by smallholder farmers in developing countries, especially concerning access to digital Agri-platforms and the eNAM model?
(Based on Joint work with Prof. Retsef Levi and Prof. Yanchong Zheng at MIT Sloan and Dr. Manoj Rajan at REMSL)
A: While digital agri-platforms hold immense potential, farmers face several challenges affecting their ability to benefit from digital agricultural platforms.
Logistics: Smallholder farmers often struggle with accessing markets due to poor transportation and infrastructure. While digital platforms increase market access, farmers must usually bring their produce to a collection point/mandi. This requirement is particularly challenging for smallholder farmers farther away from mandis. It limits their ability to sell produce efficiently and cost-effectively, impacting their profitability.
Price Transparency: Digital platforms offer greater price transparency in the market, helping farmers earn a fair price. However, farmers still need to decide when and where to sell their produce before they visit a market. Significant volatility in market prices and lack of reliable market information can still be significant barriers to improving operational decisions.
Financial Constraints: Farmers often encounter significant hurdles when accessing credit and other financial services. High interest rates, strict collateral demands, and intricate application processes compel them to sell their produce at reduced prices to traders who provide consistent credit, further undermining their financial stability and growth potential.
Digital Literacy and Access: Many farmers lack the skills to use digital platforms effectively. Additionally, inconsistent internet access and the lack of digital infrastructure in some rural areas may prevent them from utilizing these technologies to their full potential.
Q3: In your research with the Karnataka State Government, what specific digital technologies have shown promise in improving outcomes for smallholder farmers, and what are the key obstacles to their implementation?
(Based on Joint work with Prof. Retsef Levi and Prof. Yanchong Zheng at MIT Sloan and Dr. Manoj Rajan at REMSL)
A: Innovative Price Discovery Mechanisms: Digital platforms like the Unified Market Platform in Karnataka have been developed to enable better price discovery processes. These platforms allow farmers to access multiple markets and potential buyers and identify commodity prices in a transparent manner, leading to more competitive and fair pricing that benefits farmers. In our research, we experimented with a two-stage auction guided by practical operational considerations and semi-structured interviews with traders that identified behavioral factors like anticipated regret and K-level thinking affecting bidding behavior. The updated auction design was used to trade commodities worth more than $6 million USD, leading to a significant 3.6% increase in market prices for more than 10,000 smallholder farmers.
Scientific Quality Assessment: Technologies introduced to assess crop quality and health accurately can help farmers make informed decisions about crop management and sale timing. However, the implementation is often limited by the lack of local adaptation of the technologies and the training needed for farmers to use them effectively.
Warehouse-based Crop Storage Policies: Initiatives to improve crop storage and maintain quality have been linked with digital platforms to give farmers better control over when to sell their produce. This approach helps in stabilizing prices and reducing post-harvest losses. The main obstacles here include the initial cost of setting up such infrastructure and the complexity of integrating these systems with existing practices.
Price Transparency Tools: Tools that provide real-time price information across different markets can significantly enhance farmers’ negotiating power. However, the effectiveness of these tools is often reduced by inconsistent internet connectivity in rural areas and the need for continuous education for farmers on how to interpret and use this information.
Q4: Could you discuss some of the most common inefficiencies you’ve encountered in supply chains in the textile industry and how your research addresses them?
(Based on joint work with Prof. Divya Singhvi and Xinyu Zhang from NYU Stern)
A: Supervision Challenges: Because weavers work from geographically remote areas and a limited number of supervisors are expected to visit weavers physically, optimizing supervision operations is a key challenge. Because datasets from distributed artisanal supply chains is limited, there is very limited empirical evidence on the impact of supervision on productivity in artisanal supply chains. First, using supply chain data, we provide robust empirical evidence that frequent supervisor visits can play a crucial role in improving artisans’ productivity. Our results indicate that a one-day decrease in the average number of days between supervisor visits to remote weavers can increase weaving rates and monthly income by 8.6%-11.3%. We also find that (i) visits to looms with difficult-to-weave rugs and (ii) visits that are consistently scheduled have a more substantial positive impact on weavers’ productivity. To capitalize on these insights, we propose an optimization framework for scheduling supervisor visits in the supply chain using a Mixed Integer Linear Program (MILP) and show its impact in practice.
Task Assignments: Another challenge is the mismatch between the rug designs’ complexity and the weavers’ skill levels. Our field visits indicate that rug weaving is labor-intensive, and each rug takes multiple months to finish. Complex designs assigned to less experienced weavers can lead to slower production rates and incomes for weavers. We are currently working on a dynamic matching algorithm that will consider weaver experience and preferences while assigning rugs to them.
Q5: How does the utilization of machine learning models contribute to optimizing supervision, task assignment, and skill utilization among textile workers, and what impact have you observed from implementing these models?
(Based on joint work with Prof. Divya Singhvi and Xinyu Zhang from NYU Stern)
A: Machine learning models can significantly enhance efficiency in the textile industry by optimizing supervision, task assignment, and skill utilization. Our research uses state-of-the-art ML models to predict the productivity levels of different weavers based on their features (location, assigned rug, experience, etc.). Using these predictions, we then optimize the supervision schedule to target additional visits on weavers who are predicted to have low weaving rates. We implemented this ML Predict-then-optimize algorithm with our collaborator in the field, and the implementation affected almost 6,000 supervisor visits for 200 looms in Rajasthan. The empirical analysis suggests a significant gain of 16.7% in productivity and income for weavers affected by the implementation.
The post Prof. Somya Singhvi from Marshall School of Business, University of Southern California shares insights on how digital technologies empower smallholder farmers appeared first on CXOToday.com.