Delivery Order Reallocation Using Real-Time Agent and Prep Progress
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current online shopping concierge systems face challenges in optimizing the assignment of delivery tasks to agents due to uncertainties in the time taken by shoppers to collect items and delivery agents to reach stores, leading to sub-optimal assignments by the time drivers arrive, which is computationally intensive and time-consuming to resolve in real-time.
Innovation Solution
The system periodically re-allocates delivery orders among agents based on real-time data regarding the progress of order preparation and travel, locking in assignments for arriving agents and re-allocating remaining orders, thereby optimizing task assignments efficiently and quickly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the online shopping concierge service assigns different tasks to different shoppers (separating item collection and delivery), then task specialization and efficiency are improved, but coordination complexity and assignment optimization difficulty increase
Solution Approach 1:
The delivery task is segmented into two independent sub-tasks: item collection by in-store shoppers and delivery by delivery drivers. This segmentation allows each role to specialize and operate independently, improving overall productivity while the system manages coordination through separate assignment processes for each segment.
Solution Approach 2:
The online shopping concierge service acts as an intermediary system that coordinates between in-store shoppers and delivery drivers. It receives real-time data from both groups and makes dynamic assignment decisions, serving as a mediator that optimizes the handoff process without requiring direct coordination between shoppers and drivers.
2Reliability
If the system performs system-wide optimization in response to a delivery agent arriving at a store, then assignment optimality is improved, but computational time and system response time worsen
Solution Approach 1:
The system performs preliminary assignments of delivery tasks to agents before they arrive at stores, based on predicted timelines and real-time data. This preliminary action ensures that assignments are already optimized when agents arrive, eliminating the need for time-consuming re-optimization at the moment of arrival while maintaining high optimality.
Solution Approach 2:
The system performs optimization periodically based on real-time data updates from shoppers and delivery agents, rather than continuously or only at trigger events. This periodic approach balances computational efficiency with assignment optimality, updating assignments at intervals that maintain reliability without excessive computational overhead.
3Speed
If initial delivery assignments are made before shoppers complete item collection and before drivers travel to stores, then assignment speed is improved, but assignment accuracy and optimality worsen due to uncertainties
Solution Approach 1:
The assignment system is dynamic rather than static. It continuously receives real-time data from in-store shoppers about item collection progress and from delivery agents about their location and status. This dynamic information flow allows the system to adjust and re-optimize assignments in real-time, maintaining both speed and accuracy despite initial uncertainties.
Solution Approach 2:
The system implements feedback loops where real-time data from shoppers and delivery agents continuously flows back to the assignment system. This feedback enables the system to monitor actual progress against predictions and make corrective adjustments to assignments, ensuring high accuracy while maintaining rapid response times through automated real-time optimization.
Data Source
AI summary
An online shopping concierge system identifies a set of delivery orders and a set of delivery agents associated with a location. The system allocates the orders among the agents, each agent being allocated at least one order. The system obtains agent progress data describing travel progress of the agents to the location, and order preparation progress data describing progress of preparing the orders for delivery. The system periodically updates the allocation of the orders among the agents based on the agent progress data and the order preparation progress data. This involves re-allocating at least one order to a different delivery agent. When a first agent arrives at the location, the system assigns to the first agent the orders allocated to the first agent. The system then removes the first agent from the set of available delivery agents, and removes the assigned delivery orders from the set of delivery orders.