Incorporating Gen AI in logistics: optimal cloud and edge platform design for advanced driver assistance systems

Abstract

This study explores the challenges around the integration of generative artificial intelligence (Gen AI) in “cloud and edge” platform-based advanced driver assistance systems (ADAS). In this setup, vehicle-mounted edge devices provide basic driving assistance and collect data, which is then processed using cloud-based Gen AI solutions to improve driver performance. We develop an analytical model to determine the optimal balance between edge device and cloud service features, considering their complementary nature. We analyse three scenarios: a B2B market with exogenous pricing (base model), a B2C market with endogenous pricing, and a setting where firms develop only edge devices without a cloud service. We find that firm profitability and customer well-being are strategic complements, i.e., any initiative that benefits customers will also lead to higher profits for the firm. Furthermore, lower (higher) level ADAS systems are beneficial to customers at a smaller (larger) level of performance gains of the edge device. Finally, we find that customising ADAS offerings feature levels for different customer segments is crucial for the successful adoption of ADAS. Our findings provide key insights for governments and industry professionals for actionable strategies around Gen AI integration in ADAS.