AI-Powered Last-Mile Delivery: Revolutionizing OneRail with Nvidia Technology

OneRail, a delivery platform, has recently launched OmniSTAR, an AI-powered system that utilizes Nvidia technology to assist retailers, wholesalers, and distributors in making informed decisions about order deliveries. This platform evaluates various delivery options such as owned fleets, couriers, and parcel carriers, selecting the most cost-effective option that meets the required service level.

Utilizing Nvidia’s cuOpt decision optimization engine and cuDF data processing software, OmniSTAR combines OneRail’s delivery pricing and performance data. The Nvidia accelerated computing infrastructure is employed to process routing and delivery-mode calculations efficiently.

OneRail claims that the system can significantly reduce computation times, with calculations that previously took 20 minutes now being completed in under two minutes. Moreover, a week-long calculation can now be finished in about two days. This enhanced processing speed allows for real-time optimization within live delivery operations, enabling evaluation of multiple fulfillment options before assigning an order.

The broader AI systems at OneRail utilize prediction and optimization for different aspects of the delivery process. Machine-learning models estimate factors like service time, lateness risk, and first-attempt delivery success probability, which feed into the optimization systems determining order execution. OmniSTAR, specifically, compares various fulfillment modes to select the most cost-effective option based on service requirements.

Nvidia’s cuOpt, an open-source GPU-accelerated optimization library, handles route optimization and other mathematical problems. It considers factors such as vehicle costs, capacities, travel times, and operating windows when calculating routes. OmniSTAR applies cuOpt to both routing and delivery-mode selection, allowing for efficient comparison and selection of fulfillment options.

The platform also utilizes Nvidia’s cuDF, a GPU-accelerated library for tabular data processing, in combination with OneRail’s delivery data and operational models. With a dataset based on millions of deliveries across a vast network, OmniSTAR can identify cost-increasing delivery rules and assess how delivery choices impact profitability at the item level.

OmniSTAR is currently operational with selected enterprise customers, delivering tangible benefits. For instance, at US Foods, the system identified delivery configurations that were affecting margins negatively, leading to adjustments in pricing and delivery patterns. Another unnamed large tire distributor reportedly achieved significant savings using the platform.

The collaboration between OneRail and Nvidia, spanning three years before the project’s launch, involved direct engagement with Nvidia’s cuOpt engineering team and participation in the Nvidia Inception program. This partnership has resulted in innovative solutions for last-mile delivery and large-scale logistics optimization.

In conclusion, OmniSTAR’s integration of AI, Nvidia technology, and real-time optimization capabilities signifies a significant advancement in the delivery industry. With its ability to process vast amounts of data efficiently and make informed delivery decisions, the platform is poised to revolutionize the way orders are fulfilled.