Amazon.com is getting fresh with its customers in Seattle by delivering their groceries. The new service, AmazonFresh, promises next-day delivery of groceries, including fresh produce, to customers’ homes within a one-hour window. A temperature-controlled tote keeps the food fresh.

If this sounds familiar, it’s because several others have tried and failed to make variations of this business model work over the years. Whether people think it’s a good idea or a bad one, most agree materials handling technology and logistics excellence are the challenges.

“If I can make three deliveries every block, I’ll make a fortune. If I’m driving 5 miles between each delivery, I’ll lose a fortune,” says Jim Tompkins, president and CEO of Tompkins Associates, a supply chain consulting firm based in Raleigh, N.C. “You’re dealing in a business with very small margins, then you’re adding a very expensive activity. Scheduling the appointment is the challenge.”

Will Amazon’s refrigerated totes make appointments less of an obstacle?

“The cost of those totes will go up,” says Tompkins, “and there’s still a security issue: How do you get them secured unless you let people into your house? People don’t want to do that. So I think it’s a hare-brained scheme.”

Bill Bishop disagrees. He’s president of Willard Bishop Consulting and a well-known expert in the food and grocery industry. Bishop believes technology has developed enough in the last 10 years to give Amazon a fighting chance.

“The sophistication of computer scheduling, routing, planning and forecasting systems are significantly ahead of where they used to be,” he says. “Secondly, the idea of standalone operations like Peapod or Webvan has put tremendous pressure on the business model to get to scale efficiently and rapidly. Those who can do this as a joint product have some flexibility in the business model and the ramp-up to scale.”

Thirdly, he points out, materials handling technology will be a great enabler, particularly in the areas of robotics and order selection. For companies willing to make the financial and time commitments, this model can work.

“Amazon doesn’t have a stopwatch on its finances to get to critical mass, and they have a white board opportunity in terms of the physical side of assembling the orders—not to mention the refrigerated totes,” he says. “Those elements put you a lot closer to where the model could be.”