5 Burning Questions: Autonomous Vehicle Delivery

by Dave Hanson

This week’s blog really got us thinking. Not only about the ramifications of Kroger’s growing autonomous grocery delivery program, but about autonomous vehicles themselves. How safe are they? Are they actually legal everywhere? There are a lot of things we still wanted to know.

We decided to dig a little further into these issues, and it was easy to see the source of our confusion. Laws vary from state-to-state, new information regarding autonomous vehicles is always being introduced – some of it misleading – and on top of it, there are several sub-classifications within autonomous vehicles that can further convolute an already complex topic. We thought it was just manned and unmanned! Live and learn.


So, with all that being said, here are 5 Burning Questions we had regarding autonomous vehicle delivery and the answers we found.


What is the difference between manned and unmanned autonomous vehicles?

It might seem simple, but this is actually a bit of a stumbling block for many people when it comes to understanding self-driving cars. It turns out, there aren’t just two levels of autonomous vehicles here, there are actually five (six if you count no automation).

The Society of Automotive Engineers (SAE) defines the six levels of automation as follows:

Levels 0-2: This ranges from zero automation to light features like braking assistance at level 1. At level 2, you get higher-end features like steering and acceleration assistance (think cars that parallel park for you).

Level 3: This is what most people currently think of when you bring up self-driving cars. The car handles most functions, but a human driver is present and should be paying attention, as he or she may be required to pick up driving duties if conditions are less than ideal.

Level 4: This is the level where vehicles can become fully autonomous… well, to a point. The unmanned Nuro R1 vehicle Kroger intends on using is an example of Level 4. The system can work within a wide variety of pre-defined conditions, but isn’t made for all weather and road conditions. If it’s navigating neighborhood streets during normal weather and traffic conditions, there should be no problem. If it’s raining, it will probably end up being a human driver delivering your groceries.

Level 5: This is a fully automated system that can deal with all weather, traffic and lighting conditions. There are currently no Level 5 vehicles on the road in the U.S., and they aren’t expected to be on the road for another 3 or 4 years.

What are the laws regarding autonomous vehicles?

Lawmakers have been scrambling to get legislation on the books regarding autonomous vehicles. At least 36 states have now added laws governing self-driving cars, either by passing legislation through their state congress or by executive order.

The problem with this, of course, is that you have 36 states with 36 different sets of rules. This hodge-podge of different laws complicates things for automakers quite a bit and the United States’ Congress has been working to rectify the matter by creating federal legislation that will unify the laws for autonomous vehicles nationwide. An attempt to pass a bill along with the annual budget died on the table in December, but members of Congress know this is an issue that needs to be addressed. Expect to hear more on this issue this year, but safety concerns from some legislators may make passage of federal laws in 2019 difficult, as well.



How does the public feel about autonomous vehicles?

Not good. Trust is a major issue when it comes to autonomous vehicles, and Americans seem to have a deep distrust in the technology. According to a recent study by J.D. Power and the National Association of Mutual Insurance companies, 42% of Americans said they would never ride in an autonomous vehicle. Another study by AAA found 63% of drivers would be afraid to ride in a fully autonomous self-driving car.

With hundreds of billions of dollars on the line, this is a problem that manufacturers are taking very seriously. Car makers have employed teams of engineers and psychologists to work through the problem, with a team from Jaguar Land Rover recently developing a driverless pod with big googly eyes to signal the vehicle’s intentions to other drivers and pedestrians, in a recent study.

The public will likely warm to the idea as the autonomous vehicles become more common, but every time there is an accident with a self-driving car, it’s front-page news, and that has certainly damaged the public’s perception of the safety of the technology.

How safe are autonomous vehicles vs. human drivers, really?

According to U.S. safety regulators, 94% of automobile crashes can be blamed on simple human error. And with 1.3 million accident fatalities per year globally, there is obviously a lot of room for improvement in this area. If autonomous vehicles eliminate human error, one could conclude that’s 1.2 million lives potentially saved. Still, the technology seems to get a bad wrap.

According to a 2018 Bloomberg article, the vast majority of autonomous car accidents so far have been the fault of other human drivers rear-ending the driverless vehicles. The couple of incidents where the technology had a share of liability got much more attention, of course. One man was killed in Florida in 2016 when truck turned suddenly in front of his Tesla Model S, which was in self-driving mode at the time. Investigators found that both of the drivers and the autopilot were all partially at fault in this case. In a more disturbing incident that caught national headlines last year, a woman in Arizona was killed when an autonomous Uber vehicle hit her going full speed. It didn’t detect her or slow down, and investigators are still trying to piece together what exactly happened in this case. In the meantime, Arizona has suspended Uber’s autonomous vehicle program until definitive conclusions can be reached.

But these accidents are far more the exception than the rule. Autonomous vehicle crashes are extremely rare overall. With human error accounting for 19 out of every 20 crashes in the U.S., it doesn’t take a rocket scientist to figure out that eliminating human error eliminates a lot of crashes. A recent study by Morgan Stanley estimated that autonomous cars could save the U.S. $563 billion per year in costly accidents, simply by eliminating human error, and $1.3 trillion overall with all savings factors considered, such as traffic, fuel, congestion and productivity savings. Getting bullish, it could even stretch up to $2.2 trillion. So maybe these autonomous cars are a bit safer than many of us presume.

If this program is successful, what will the impact be on jobs?

There are simply no two ways about it. Unmanned vehicles like the Nuro R1 require fewer employees. The switch from manned to unmanned delivery vehicles will cut several potential jobs from the rapidly growing service of grocery delivery at any location that makes the swap, and the U.S. is already strained for unskilled labor jobs, as it is.

According to a frequently cited 2013 Oxford study, up to 47% of jobs in the U.S. are at high risk of automation in the next few decades. A more recent report from McKinsey estimated that up to 800 million jobs could be wiped out due to automation by 2030. However, there is little actual consensus among experts as to how many jobs will actually be lost.

What we can say is that a quick search for grocery delivery driver jobs on Snagajob.com returned some 68,000 listings. Ziprecruiter.com also retuned a similar number, once we started combining the salary ranges. If autonomous delivery grows, these listings will surely plummet.

Looking at Kroger specifically, who owns nearly 2,800 stores across the US., it is easy to see how these numbers could really add up. A fleet of 10 unmanned vehicles per store would eliminate 28,000 potential delivery jobs. Though in fairness, there would likely need to be some programmers and maintenance people added to service the fleet. Perhaps that would add 8,000-10,000 jobs back into the fold. Considering the rapid expansion of delivery as a service, one could look at this as a net positive, relative to an era where grocery delivery jobs of any kind barely existed just a short time ago. It’s all a matter of perspective, and one can hardly fault Kroger for trying to run an efficient business.

Furthermore, as we mentioned earlier, autonomous vehicle technology isn’t at the point where it works across all types of weather and driving conditions. The Nuro R1 is a Level 4 vehicle. Level 5 vehicles are still a few years out. So expect Kroger to continue hiring drivers for their rapidly growing delivery business in the meantime.


Join the Discussion

What kind of burning questions do you have about autonomous vehicle delivery? What concerns do you have? Join our discussion on LinkedIn over the next few days for polls, discussions and more on this hot button issue!