The future of “last-mile” delivery of e-commerce orders is one of flying drones, robots and driverless vehicles rushing product to our doorsteps. Or so the futurists would have us believe.

As with most technological marvels, however, reality must at some point step in to temper the dream.

The desire to embrace technology is understandable. With the growth of e-commerce — accelerated by the coronavirus pandemic — comes the challenge of delivering millions of packages in congested urban environments, remote rural areas and sprawling suburbs. It’s a task that can’t possibly be carried out by humans and conventional vehicles alone.

So we look to “smart” machines that can get the job done cheaply and efficiently, with the ability to conform to any residential, commercial or industrial setting. The most popular of those options is the flying drone, which has been under aggressive development in recent years by such tech giants as Amazon and Google, and delivery services such as UPS.

Most modern drone designs take the form of four to six rotors with vertical takeoff and landing capability. The package is secured to the underside of the unit and lowered to the ground via cable. Variants are already in limited operation around the world. The operator Flytrex, for one, has carried out numerous successful runs in Iceland, and is seeking authorization to expand operations to North America and elsewhere.

Authorization is, of course, one of the chief stumbling blocks to full deployment of drones. The U.S. Federal Aviation Administration and corresponding agencies in other countries have been mulling over how to regulate these small craft, with their potential for causing injuries to people, damage to power lines and, especially, interference with aviation. A single errant drone, drifting too close to the flight patterns of a major airport, can have devastating consequences.

Issues of practicality, while more mundane in nature, further restrict the widespread use of drones, especially in cities. How can they efficiently deliver to the doorsteps of buildings with multiple units? Do they need to be fitted out with rooftop landing areas that serve all residents? And what about the sound generated by those whirring rotors? Won’t they amp up already high levels of noise pollution?

With such barriers in mind, the most promising near-term use of flying drones might be in remote areas such as countries within Africa, where other delivery methods aren’t available, says Panagiotis Tsiotras, professor at the Guggenheim School of Aerospace Engineering at the Georgia Institute of Technology and Institute of Electrical and Electronics Engineers (IEEE) fellow. They can be a crucial means of supplying essential items such as food and medicines.

In many such cases, the drones would have to be able to fly over long distances. But endurance remains an issue, notes Tsiotras, with batteries often lasting no longer than around 15 minutes. “If we solve that problem,” he says, “it will be a much more mature technology.”

Finally, there’s the issue of economic viability. What’s the cost of using a expensive, sophisticated drone to deliver a single pizza? How many trips must it make to justify investment in the technology? Would it ever reach that point?

So have delivery drones been overhyped? Without question, but so has every major technological innovation in its early stages of development. Many problems remain to be ironed out, but Tsiotras believes drones already make business sense in certain commercial applications. In theory, they could operate effectively over short distances, with multiple units launched from trucks positioned at the edge of busy neighborhoods. The suburbs also present an attractive environment, especially where private homes have yards in which packages could be safely dropped.

Driverless trucks and vans have also been touted as a solution to the last-mile conundrum, although they come with their own set of questions about safety, reliability and public acceptance. Tsiotras believes the technology will ultimately prove more feasible on major highways than in heavily populated cities. They could provide an answer to the persistent commercial long-haul driver shortage, which is only expected to worsen in the years ahead. At least a dozen companies are seriously testing the technology, he says, with limited pilots underway in places such as Arizona.

A sky full of drones? Highways crowded with driverless big rigs? Robots skittering along sidewalks on their way to delivering packages? With any new technology, there’s always a lag from idea to viability, says Tsiotras. And real-world applications are likely to differ markedly from the original dream.

In the end, practicality prevails. “The good ideas will survive,” Tsiotras says, “and bad ones will die out.”