Can AI Revolutionize E-Commerce Last-Mile Delivery? A Driver’s Perspective

In the race to improve last-mile delivery, artificial intelligence (AI) is often touted as the ultimate tool for efficiency. But a recent webinar on the topic raised a thought-provoking question: How committed are delivery drivers to following AI-generated instructions?

To the experienced delivery drivers, the answer is obvious. For others, it isn’t as clear-cut.

The Human Advantage

Let’s consider the reality:

• New Drivers: It takes 4-6 weeks for a new driver to know their route (typically 110-160 streets) better than a navigation system, on a “true address” level.

• Experienced Drivers: For drivers switching to a new route, it’s about 2-3 weeks to adapt.

• Customer familiarity: Over about 6-8 months, drivers begin to build a friendly relation  with their customers, improving their efficiency and service quality substantially.

AI, on the other hand, relies on analyzing historical digital generated data: GPS coordinates, delivery times, and other statistics generated by algorithms. But there is a catch:

The Key Questions About AI in Last-Mile Delivery

1. Data Quality:Who generated the data? Was it by experienced drivers or those still learning the trade? Was the data collected correctly, or was it sloppy or even manipulated under pressure?

2. Adapting to Change:AI struggles with dynamic factors like changing customer behavior, route adjustments due to staffing shortages, or unpredictable environmental challenges.

3. Overlooked Realities:Human and environmental factors are critical. For instance, UPS’s Orion system faced significant hurdles, as highlighted by their own drivers. Similarly, Waymo’s autonomous taxis caused a parking lot traffic jam because the system overlooked a simple logistical detail.

Why the Human Element Matters

The common argument is that AI can “learn.” But AI’s success hinges on collaboration with the people engaging in the last-mile process: from the customers placing the orders, to the drivers delivering them, and the recipients receiving them.

Without actively involving delivery drivers in developing AI systems, these tools will fall short of their potential. e-Commerce delivery is a human-centric process, and ignoring that reality limits the effectiveness of even the most advanced technologies.

Moving Forward

AI holds promise, but it’s not a standalone solution. By integrating drivers’ insights and addressing real-world complexities, AI can enhance efficiency while maintaining the human touch that makes last-mile delivery work.

What’s your take? Is AI ready to truly transform the last-mile delivery landscape?