International Review of Business, Trade, and Economics
Algorithmic Logistics and Labor Economics in the On-Demand delivery Ecosystem: A Socio-Technical Analysis with a Proposal for Rider-Centric Intelligence
Abstract
Narayana Sanghea Panchumarthy
Food delivery has, in less than a decade, moved from a neighbourhood trade conducted by phone to a globally orchestrated, platform-mediated logistics industry. What looks to a customer like a thirty-minute ride from a kitchen to a doorstep is, underneath, a chain of sub-second decisions taken by forecasting engines, dispatch heuristics, sensor- fusion models and pricing controllers. This paper offers a structured reading of that stack and of the labour economy it produces. I trace the methodological evolution of demand forecasting from k-nearest-neighbour regressions to recurrent and probabilistic deep networks, examine dispatch and matching engines through Deliveroo's Frank algorithm and Uber Eats' triple-batching logic, and explain how sensor fusion - accelerometers, gyroscopes, conditional random fields converts a noisy GPS trace into a fine-grained model of the courier's working day. I then turn to the economic and social consequences. Drawing on Berkeley Labor Center net-pay data, the Stanford rideshare gender-gap study, and recent reporting on the Bologna Frank ruling, the New York City Department of Consumer and Worker Protection minimum-pay rule, the EU Platform Work Directive 2024/2831 and China's 2027 algorithm-bargaining mandate, the paper argues that the gig delivery system has now passed into a 'hyper-algorithmic' phase whose efficiency cannot be separated from its labour-market effects. As a constructive contribution, I outline FindMyRider, a rider-centric intelligence platform combining demand forecasting, density-aware heatmap visualisation and explainable, contestable decision logs -intended as a research blueprint, not as a commercial system. The paper closes with limitations, ethical reflections and an agenda for further work on retrieval-augmented, human-in-the-loop logistics analytics.