Artificial Intelligence in Collaborative Data Sharing for Intra-Company Logistics
forms the essential prerequisite for the horizontal expansion of AI capabilities to adjacent scenarios and data sources. By generating reusable, high-quality, and trusted data assets, this foundation advances sustainability objectives and supports the emergence of smart operations and adaptive supply chain processes.
4 AI-SUPPORTED ANALYTICS AND OPTIMIZATION
Complementing the robust data provisioning and operational integration layer, the second AI dimension, AI-supported analytics and optimization, unlocks advanced decision support and process improvement capabilities within the Testbed environment. A range of AI approaches are applicable and, in part, already prototyped: predictive analytics to anticipate logistical disruptions or asset failures, anomaly detection algorithms operating on high-frequency sensor streams to identify deviations in real time, and optimization routines that enhance routing, scheduling, and resource allocation under dynamic constraints.
These capabilities align closely with the broader evidence on AI-driven intelligent automation in supply chains. Shamsuddoha et al. [ 5 ] highlight how such technologies deliver concrete performance gains in logistics optimization, demand forecasting, warehouse and inventory efficiency, real-time operational decision-making, and proactive risk mitigation. When embedded in collaborative data-sharing architectures, these analytics functions amplify the value of shared data into actionable, cross-partner insights that improve both local efficiency and ecosystemwide coordination.
Example
In the ICL Testbed, the shared IoT data enables precise capacity planning for daily operations. Planners can determine how many vehicles are required for timely truck unloading at the receiving area and how many additional units are needed for subsequent transport to production lines or storage zones. Without reliable, near-real-time visibility into actual utilization patterns, such planning often relies on experience or static rules, leading to either idle assets or bottlenecks during peak periods. AI-supported analytics add a data-driven layer to this process. Clustering techniques applied to historic operational data reveal recurring patterns in workload distribution across shifts, zones and task types, allowing planners to anticipate required fleet sizes more accurately. These insights are further refined through direct feedback from the planning role: when dispatchers adjust AI-generated recommendations the system could incorporate these corrections, progressively improving the quality of future capacity forecasts and closing the loop between automated analysis and human expertise.
Empirical work by Culot et al. [ 4 ] further substantiates the performance implications of AI deployment in supply chain management contexts, underscoring contributions to operational outcomes through more efficient utilization of resources and data assets. Importantly, the enriched models, patterns, and decision rules generated through these analytics activities do not remain confined to the core Track and Trace use case. Instead, they create the cognitive and
EDM Association – Journal of Innovation 131