*

Amazon ( AMZN ) to use AI for advanced warehouse robots

*

Generative AI to enhance delivery maps for Amazon ( AMZN ) drivers

*

AI to optimize product stocking, reduce carbon emissions

(Adds executive comments throughout)

By Greg Bensinger

SUNNYVALE, Calif., June 4 (Reuters) - Amazon ( AMZN )

wants customers to know that artificial intelligence is not just

for writing college essays.

In a series of announcements Wednesday, Amazon ( AMZN ) demonstrated

how stockroom robots, delivery people and its sprawling

warehouses will all benefit from a hefty dose of artificial

intelligence, speeding packages to customer doorsteps.

The company said it is forming a new group at its Lab126

device unit focused on creating warehouse robots that will

perform multiple tasks when prompted, a significant advance over

today's robots that typically are designed for a singular job.

Using so-called agentic AI, these robots will be able to

unload trailers and then retrieve parts for repairs, according

to Amazon ( AMZN ).

"For our customers, it's, of course, faster delivery," said

the unit's leader, Yesh Dattatreya, a robotics scientist, at an

event at Amazon's ( AMZN ) Silicon Valley Lab126 hardware device lab. He

said the robots could be critical during times of heavy demand,

like around the holidays, for things like lifting heavy objects

in confined spaces.

The new AI would also help the company minimize waste and

cut carbon emissions, Amazon ( AMZN ) said.

Agentic AI has become one of the hot investment areas with

technologists promising software that can make decisions and act

upon them without any additional input from users. Such software

is meant to help automate everyday tasks like scheduling.

"We're creating systems that can hear, understand and act on

natural language commands, turning warehouse robots into

flexible, multi-talented assistants," Amazon ( AMZN ) said in a statement

prior to the lab event.

Dattatreya said decisions like what the robots would look

like, how many would be deployed or when had yet to be

determined.

NAVIGATING OBSTACLES AT DELIVERY POINTS

Amazon ( AMZN ) is also using generative AI to create more advanced

maps for its delivery drivers, so that they can more efficiently

deliver packages. The specialized AI will provide Amazon ( AMZN ) fine

detail on building shapes, as well as obstacles and anything

else they may need to navigate for a package drop-off.

"This innovation is making it easier for Amazon ( AMZN ) drivers to

find the right delivery spot, especially in tricky places like

big office complexes," Amazon ( AMZN ) said.

That technology could be critical to specialized eyeglasses

Amazon ( AMZN ) is developing for delivery drivers that Reuters reported

exclusively last year. The company hopes to outfit drivers with

screen-embedded glasses that free their hands from GPS devices

and give them turn-by-turn directions while driving, as well as

while carrying packages at their destination.

Viraj Chatterjee, a vice president in Amazon's ( AMZN ) Geospatial

unit, said in an interview the technology could potentially be

used in the eyeglasses, but that the hardware was far from being

perfected. It marked Amazon's ( AMZN ) first public acknowledgement of

the eyeglass project.

He said delivery drivers in the U.S. were already using the

maps daily. The software has been particularly useful for large

apartment complexes and housing developments, he said.

Chatterjee said Amazon's ( AMZN ) delivery people, who are largely

contractors, are not required to use the software. Some gig

companies have faced legal challenges over whether they are

asserting too much control over their contracted workers through

mapping and other software.

Amazon ( AMZN ) also said AI will help it more efficiently predict

what products customers will need and where to improve its same

day delivery operations, with the software considering factors

such as price, convenience, weather and sales events, like Prime

Day.

"It allows us to sell a different set of books in Boston

than we would in Boise, and cater to different tastes really,

really efficiently across the communities that we serve," said

Nathan Smith, director of demand forecasting for Amazon's ( AMZN ) supply

chain optimization technologies unit.