Amazon has rolled out several new robotic technologies in its warehouses, claiming that these innovations will enhance productivity and minimize injuries among workers.
Two notable robotic systems introduced are the robotic arms named Robin and Cardinal, designed to lift packages weighing up to 50 pounds. Another robot, known as Sparrow, specializes in transferring items between bins and containers.
Additionally, Proteus is an autonomous mobile robot that navigates the warehouse floor to transport carts, while Digit, a humanoid robot that walks on two legs, is currently being trialed to assist in moving empty totes. There’s also Sequoia, a storage solution that organizes totes for employees, allowing them to retrieve items without the need for bending down or reaching excessively.
According to Amazon, Robin is already operational in numerous warehouses, whereas the other robots are still undergoing testing or have not been widely implemented yet. The company asserts that it has noticed positive outcomes, including faster order fulfillment and reduced repetitive tasks for employees. Nevertheless, advancements in automation might necessitate retraining for staff members if their current jobs become redundant.
In October, Amazon hosted an event at a warehouse in Nashville, Tennessee, where some of these robots were incorporated. During this event, Julie Mitchell, the director overseeing Amazon’s robotic sortation technologies, discussed the company’s future plans. Below are excerpts from her conversation, which have been edited for conciseness and clarity.
Q: What is the typical timeline for deploying new robotic technology?
A: Our journey with robotics has spanned several years, supported by over a decade of foundational work. We began integrating Cardinal and Proteus into this facility in November 2022, focusing on optimizing packaging and order movement. Now, less than two years later, we are processing about 70% of outgoing items through this robotic system.
Q: So it takes about two years?
A: Yes, our development cycle is generally “build, test, and scale,” which currently lasts around two years.
Q: What challenges arise when creating robots that can physically handle products?
A: There is a significant challenge due to the vast variety of items we manage. We rely heavily on data, and our first prototype is placed in a real-world environment to evaluate its performance, allowing us to identify and rectify failures quickly. For instance, after launching the Robin robotics arm—a package manipulation tool—we achieved 3 billion item picks. Our ability to rapidly innovate, collect data, and iterate has accelerated our progress.
The challenges boil down to three main components: sensing the environment, planning movement, and executing tasks. Currently, these operations are managed separately, but advancements in artificial intelligence will enable us to streamline this process. We are on the cusp of utilizing generative AI to significantly enhance the functionality of our robotics.
Q: What considerations do you have regarding automation’s effect on Amazon’s workforce during this development?
A: The deployment of this technology is creating new roles that require individuals to gain additional skills. Importantly, these skills are not overly technical; they do not necessitate advanced degrees. We’ve engineered these systems to be user-friendly, enabling employees to be trained on-site for maintenance roles.
Our objective is to foster a workforce with higher skill levels, which comes with increased pay compared to entry-level positions. Collaborating with MIT has allowed us to focus on what is most beneficial for our employees as we roll out these technologies.
Q: Are there obstacles to integrating these robots into your warehouses?
A: So far, we haven’t faced major hurdles with adoption. However, these systems are complex, and real-world conditions can introduce complications. For instance, inclement weather in the Southeastern region can disrupt our logistics. By analyzing data from our robotic systems, I can track delays, especially during bad weather when shipments can be impacted.
Delays in truck arrivals or departures lead to unexpected bottlenecks in our warehouses. Thus, effective communication between our robotic systems and human staff is key to navigating such logistical challenges. The goal is not for robots to fully take over but to create a collaborative environment in which humans and robotics work together cohesively to ensure efficient product shipping.