Latest news with #AmazonRobotics
Yahoo
25-07-2025
- Business
- Yahoo
Amazon Now Employs Almost As Many Robots As People
Amazon has deployed its one-millionth robot. The e-commerce giant uses robots in its warehouses and facilities to move products around, handle packages on a conveyor belt and assist human workers. Robot No. 1 million is now active at an Amazon facility in Japan, one of about 300 facilities where Amazon has started using robots, the company announced in a blog post. More from Sourcing Journal Senate Axes AI Moratorium from 'Big, Beautiful Bill' Byte-Sized AI: Walmart Announces New Employee Tech; Startups Bag Funding for Virtual Try-On Salesforce: Consumers' Product Discovery Expectations Are Shifting Ahead of Holiday The company first started using robots in the warehouse in 2012, when it acquired Kiva Systems—which has since been renamed Amazon Robotics—for $775 million. Since then, it has iterated on its technology strategies, releasing robots that have a variety of form factors and use cases. Its latest and greatest is Proteus, an autonomous mobile robot (AMR) that can move through the warehouse without human intervention. That it has reached one million deployed robots puts Amazon's technologically based workforce close to the number of human employees it has. The company is the U.S.'s second-largest private employer, and according to the Wall Street Journal, it employs more than 1.5 million people. Now that those people—most of whom work in warehouses and fulfillment centers around the world—are moving alongside several different types of robots in the warehouses, the company has also unveiled a robot-wrangling system called DeepFleet, which it designed to coordinate the movement of robots within the warehouse. Because Amazon deploys robots with multiple form factors in some of its facilities, it's possible for them to block one another's paths or force each other to take less-efficient routes throughout the warehouse. Scott Dresser, vice president of Amazon Robotics, said DeepFleet will reduce robots' travel times by about 10 percent. 'Think of DeepFleet as an intelligent traffic management system for a city filled with cars moving through congested streets,' Dresser wrote in a blog post detailing the upgrade. 'Just as a smart traffic system could reduce wait times and create better routes for drivers, DeepFleet coordinates our robots' movements to optimize how they navigate our fulfillment centers. This means less congestion, more efficient paths, and faster processing of customer orders.' DeepFleet, like many of today's AI-based technologies, will only become more effective as it ingests further data and warehouse scenarios in real time, Dresser said. While Dresser's blog post did not make it clear whether the two projects were interconnected, researchers at the Massachusetts Institute of Technology (MIT) announced last year that they had been using deep-learning models to segment warehouse robots in an effort to control traffic and curtail accidents—and MIT disclosed that Amazon supported that project at the time. According to Amazon, robots have a role in handling three-quarters of all customer orders globally. The company contends that in the process of building out its robotic fleets it has also upskilled more than 700,000 employees in an effort to prepare them for the future of work, which indubitably involves AI systems, robots and more. Still, in a memo last month, CEO Andy Jassy noted that the company will likely shed jobs because of the further proliferation of AI—and, in particular, generative AI. The warehouse isn't the only terrain Amazon's testing emerging technologies in; it's also been bullish on bringing drones—and more recently, robots—to delivery sites. Amazon has Prime Air sites in Texas and Arizona, where it works to deliver small parcels to consumers' homes rapidly. It has beyond visual line of sight (BVLOS) permissions from the Federal Aviation Administration (FAA) for its MK30 drones, which means the drones can travel autonomously beyond what its operator can see. Last month, The Information reported that Amazon had started building a 'humanoid park' to test humanoid robots trained by AI to deliver parcels to consumers' homes. The idea is that, eventually, a humanoid would ride in the back of an Amazon truck and pop out to leave a package at a delivery site, easing the burden on the human driver. Already, Amazon has been creeping up on competitors in the logistics sphere; added robotics and autonomous transport capabilities could continue pushing its ranking up. New data from the Pitney Bowes Parcel Shipping Report shows that Amazon Logistics made greater gains than its competitors at UPS, USPS and FedEx in 2024. USPS saw a 3.2-percent uptick in parcel volume, while Amazon Logistics saw a 7.3-percent increase year on year. FedEx, meanwhile, saw its volume slip by 3.6 percent, and UPS held steady with 1.7-percent year-on-year growth. Shemin Nurmohamed, executive vice president and president of sending technology solutions at Pitney Bowes, said such a trend is indicative of a shifting third-party logistics market. 'Since Pitney Bowes began tracking shipments a decade ago, the parcel market has been dominated by FedEx, UPS and USPS. We are witnessing a turning of the tide, evidenced by the nearly 40% volume growth in the five-year CAGR of 'Other' carriers. This disruption presents a unique opportunity for businesses to take advantage of competitive pricing,' Nurmohamed said in a statement.
Bloomberg
08-07-2025
- Business
- Bloomberg
S&P 500 IS ROUGHLY FLAT
Bloomberg Television brings you the latest news and analysis leading up to the final minutes and seconds before and after the closing bell on Wall Street. Today's guests are Abby Yoder J.P. Morgan Private Bank, Matthew Luzzetti Deutsche Bank, Scott Dresser Amazon Robotics, Ed Ludlow, Bloomberg Tech, Jay Woods Freedom Capital Markets, Sara Hsu University of Tennessee at Knoxville, Matt Maley Miller Tabak, Ralph Schlosstein Evercore and Jason Warner Poolside. (Source: Bloomberg)
Bloomberg
08-07-2025
- Business
- Bloomberg
How Robots Are Helping Amazon Deliver on Prime Day
Amazon Robotics Vice President Scott Dresser explains how robots are helping Amazon employees deliver on Prime Day. He speaks on "Bloomberg The Close." (Source: Bloomberg)
Forbes
07-07-2025
- Business
- Forbes
The People Have Spoken About Trump's AI Plan. Will Washington Listen?
Tech leaders urge light-touch regulation as public calls for accountability grow — a divide at the ... More heart of the U.S. AI Action Plan, due out any day now. U.S. Senate Commerce Committee hearing on AI on May 8, 2025. (Photo by BRENDAN SMIALOWSKI) This article was written by Paulo Carvão, with Mizuki Yashiro, a sophomore studying Economics and Government at Harvard, serving as the Director of Strategy at Harvard Venture Capital Group and a data science automation intern at Ategrity Specialty Insurance, and Shaurya Jeloka, a sophomore studying computer science and economics at Harvard and interning as a software engineer at Amazon Robotics. The U.S. Artificial Intelligence Action Plan is due any day now, and the stakes couldn't be higher. The Trump administration asked the public earlier this year to help shape the plan. Over 10,000 responses poured in from tech giants, startups, venture capitalists, academics, nonprofit leaders and everyday citizens. What emerged from this unprecedented consultation is not just a collection of comments. It's a revealing portrait of the tensions shaping America's AI debate. The country is divided, not only between industry and civil society, but within the tech sector itself. If the U.S. is to lead responsibly in AI, federal policymakers must look beyond industry talking points and confront the deeper-value conflicts that these responses lay bare. Our team analyzed the full set of public comments using a combination of machine learning and qualitative review. We grouped responses into six distinct 'AI worldviews,' ranging from accelerationists advocating rapid, deregulated deployment to public interest advocates prioritizing equity and democratic safeguards. We also classified submitters by sector: big tech, small tech (including VCs) and civil society. The result offers a more structured picture of America's AI discourse and a clearer understanding of where consensus ends and conflict begins. Industry and civil society are polar opposites: 78% of industry actors are accelerationists or national security hawks, while close to 75% of civil society respondents focus on public interest and responsible AI advocacy. The Innovation vs. Governance: A Fault Line Tech companies overwhelmingly support U.S. global leadership in AI and warn against a fragmented regulatory landscape. OpenAI called on the federal government to preempt the 'patchwork of regulations' that risk 'undermining America's leadership position.' Meta warned that diverging rules 'could impede innovation and investment.' Leading VCs, including Andreessen Horowitz and True Ventures, echoed these concerns, cautioning against 'preemptively burdening developers with onerous requirements' and pushing for a 'light-touch' federal framework to protect early-stage startups from compliance burdens. The House included a controversial provision in Trump's budget bill that would have imposed a 10-year ban on state-level AI regulation, but the Senate struck it down Tuesday, sparking renewed debate. Yet these voices are far from unified. Traditional enterprise firms like Microsoft and IBM adopt a more measured stance, pairing calls for innovation with proposals for voluntary standards, documentation and public-private partnerships. In contrast, frontier labs and VCs resist binding rules unless clear harms have already materialized. Meanwhile, civil society groups, ranging from the Electronic Frontier Foundation to the Leadership Conference on Civil and Human Rights, argue that those harms are not hypothetical, but are here now. Biased hiring algorithms, surveillance creep in policing, and opaque decision systems in healthcare and housing have already caused real damage. These organizations support enforceable audits, copyright protections, community oversight and redress mechanisms. Their vision of 'AI safety' is grounded not in national competitiveness, but in civil rights and systemic accountability. Shared Priorities, Divergent Principles Despite philosophical divides, there is some common ground. Nearly all industry actors agree on the need for federal investment in AI infrastructure, energy, compute clusters and workforce development. Microsoft has committed $50 billion to U.S. AI infrastructure; Anthropic warned that powering a single model might soon require five gigawatts of electricity. Industry wants government support to scale AI systems and do it fast. But when it comes to accountability, consensus collapses. Industry prefers internal testing and voluntary guidelines. Civil society demands external scrutiny and binding oversight. Even the very definition of "safety" differs. For tech companies, it's a technical challenge; for civil society, it's a question of power, rights and trust. Why This Matters for the Action Plan Policymakers face a strategic choice. They can lean into the innovation-at-all-costs agenda championed by accelerationist voices. Or they can take seriously the concerns about democratic erosion, labor dislocation and social harms raised by civil society. But this isn't a binary choice. Our findings suggest a path forward: a governance model that promotes innovation while embedding accountability. This will require more than voluntary commitments. It demands federal leadership to harmonize rules, incentivize best practices, and protect the public interest. Congress has a central role to play. Litigation and antitrust cases may offer remedies for past harms, but they are ill-equipped to prevent new ones. Proactive tools, including sector-specific regulation, dynamic governance frameworks and public participation are needed to build guardrails before disaster strikes. Crucially, the government must also resist the temptation to treat 'the tech sector' as a monolith. Our analysis shows that big tech includes both risk-conscious institutional players and aggressive frontier labs. Small tech spans open-source champions, privacy hawks and compliance minimalists. Civil society encompasses not only activists, but also major non-tech corporations such as JPMorgan Chase and Johnson & Johnson, whose AI priorities often bridge commercial and public interest values. Bridging the Divide There is no perfect formula for balancing speed and safety. But failing to bridge the value divide between industry and civil society risks eroding public trust in AI altogether. The public is skeptical, and rightfully so. In hundreds of comments, individuals voiced concerns about job loss, copyright theft, disinformation and surveillance. They didn't offer policy blueprints; instead, they demanded something more essential: accountability. If the U.S. wants to lead in AI, it must lead not just in model performance, it needs to lead in model governance. That means designing a system where all stakeholders, not just the largest companies, have a seat at the table. The Action Plan must reflect the complexity of the moment and should not merely echo the priorities of the powerful. The people have spoken. The challenge now is whether Washington will listen — not just to those who build the future, but to those who must live in it.
Forbes
07-07-2025
- Business
- Forbes
Amazon's Millionth Warehouse Robot Is Here And It's Getting Smarter
An Amazon employee checks the robotic warehouse at Amazon's Robotic Fulfillment Centre in Sutton ... More Coldfield, England. Amazon Robotics announced last week that it just hit a major milestone with over one million warehouse robots deployed. In addition to the announcement, the company unveiled DeepFleet, a new AI foundation model that acts like a traffic controller for its machines, helping them move more efficiently avoid bottlenecks. Amazon CEO Andy Jassy recently tweeted a simple question: 'How can we help operations employees access inventory more efficiently?' The company says that's the north star for its robotics strategy. These machines aren't about replacing people, they're about speeding things up. From a Handful to a Million Robots Amazon started testing shelf-moving robots back in 2012. Since then, the use of warehouse-floor robots has become central to how Amazon fills orders, with over a million robots now operating across more than 300 warehouses. These robots are now the workhorses of Amazon's fulfillment centers. The millionth unit came off the production line in Japan, marking the shift from simple conveyors to fully automated helpers. These bots aren't just flashy, but they do heavy lifting, literally. The robots are squat robotic carts the size of a footstool. They slide under a shelving tower, lift the whole rack, and bring it to a worker who plucks the ordered item. When the shelf is empty, the robot puts it back and hunts for the next pick. Every trip it makes shaves a little walking time from a human packer's shift and pushes an order closer to getting out the door. It's all about cutting down the time it takes to get an order from shelf to shipping box. DeepFleet: Real-Time Routing for Robots The newly announced DeepFleet system is built on Amazon's AWS infrastructure and uses live warehouse data to reroute robots on the fly. It looks for jams or bottlenecks and finds better paths, helping robots avoid wasting time in the warehouse equivalent of a traffic jam. 'Think of DeepFleet as an intelligent traffic management system for a city filled with cars moving through congested streets,' says Scott Dresser, VP of Amazon Robotics. 'DeepFleet coordinates our robots' movements to optimize how they navigate our fulfillment centers. This means less congestion, more efficient paths, and faster processing of customer orders.' The new AI model learns on the fly to steer bots around slow spots, cutting their travel time by about 10 percent and speeding up customer orders. That smarter routing means fewer delays and more predictable delivery times for customers. According to Amazon: More Tech Jobs, Not Fewer In Shreveport, Louisiana, Amazon opened a new fulfillment center last year with its latest generation of robots. According to the company, those robots didn't replace jobs, they created new ones. 'Advanced robotics require 30% more employees in reliability, maintenance, and engineering roles,' Dresser said. New roles range from motor repair techs to data analysts. Repair techs keep the bots rolling, and the analysts use data insights to track robot movement patterns to spot delays. It's not just about operating machines for these new roles, it's about maintaining and improving them. Since 2019, more than 700,000 Amazon workers have joined programs to build up their tech skills. Whether those workers stay at Amazon or move on, the company says the experience prepares them for jobs in an increasingly automated economy. The Bigger Automation Picture Amazon's not alone in using bots to automate the warehouse. U.K.-based Ocado uses thousands of compact bots in its grocery warehouses. The bots zip across grid-like platforms, picking items with near-perfect accuracy. The system can pack a customer's order in seconds. Walmart is working with Symbotic to bring robot-powered sorting to 42 of its distribution centers by 2030. Their CasePick robots can sort cases in under a minute, helping speed up restocking in stores. The trend is clear: companies are investing in automation not just to move faster, but to rethink how and where human workers fit into the process. What's Next for Amazon Robotics Amazon plans to use data from DeepFleet to cut down energy use inside its warehouses. The company is also experimenting with smaller, more nimble 'micro-fulfillment' centers closer to big cities. That could mean faster delivery and a smaller carbon footprint. Eventually, Amazon says AI won't just direct traffic. It'll help decide what products to stock and when, turning warehouses into systems that adjust themselves in real time. 'This is just the beginning. As DeepFleet learns from more data, it will continue to get smarter, driving deeper efficiencies, unlocking more selection closer to customers, and reimagining what's possible in robotic logistics,' Dresser says. The big unknown? How people will fit into a world where machines not only move the shelves but make the decisions too.
