AMMR stands for autonomous mobile manipulation robot. These robots combine autonomous navigation with autonomous grasping. With a wheeled mobile base and visual sensors that allow them to sense where they are in their environment, AMMRs are able to navigate from point A to B with the ability to “see”. The robots also are equipped with an arm and end-effector which allows them to physically pick up different items.
IAM Robotics created the first and only commercially viable AMMR, the Swift Product Suite, which was launched in 2016. Today, Swift is deployed in warehouses and distribution centers to pick and move products for order fulfillment.
Goals of Autonomous Mobile Manipulation
The goal of autonomous mobile manipulation as defined by the IEEE Robotics & Automation Society is to execute complex manipulation tasks in an unstructured and dynamic environment where cooperation with humans may be required. To achieve this goal, scientists and engineers — including the roboticists at IAM Robotics — are addressing several technical challenges:
- Generality – How to perform a variety of tasks, gain new skills, and apply these skills in novel situations; how to adapt continually and improve performance
- High Dimensional State Space – How to use many actuators and sensors to allow for greater high-dimensional state spaces
- Uncertainty – How to address problems that arise due to the uncertainty of sensing and actuation
- System Complexity – How to integrate numerous hardware components for sensing, manipulation, and locomotion to orchestrate algorithmic capabilities in perception, manipulation, learning, controlling, planning, etc.
How IAM Robotics Approaches AMMR Development
While addressing the challenges described above, the roboticists at IAM Robotics have identified four key development principles that guide their work:
- Robustness – Industrial automation requires a solid, durable, stable platform, and the technology must provide expansive features and capabilities to provide customers with flexibility. Robustness is the foundation from which the rest of the IAM Robotics principles stem from.
- Sensing – The technology requires the most advanced state-of-the-art computer vision combined with an onboard data network to provide robots with the capability to make decisions extremely fast resulting in maximum throughput.
- Perception – The technology requires perception software/algorithms that allow robots to identify and manipulate objects and to navigate through environments designed for humans.
- Manipulation – The robots must be constructed to handle objects of various sizes and weights and with high levels of dexterity.
AMMR vs. AMR
An autonomous mobile robot, or AMR, has many of the same attributes as an AMMR but lacks the property of manipulation (i.e., grasping, picking). Warehouse and distribution center operators use AMRs as mobile carts, transporting totes or bins of goods. Products are placed onto and retrieved from the AMR by human workers or other automation. In order fulfillment applications, AMRs typically either follow picking workers, pushing them to pick faster, or lead picking workers, guiding them to move faster to the right slotting locations and allowing them to work in a more hands-free environment.
In this application, AMRs save time by somewhat reducing the number of steps that a worker has to take during a shift. But, AMRs require a human worker on either end of the workflow—taking the products off of the racks, putting them onto the carts, and then taking the product off the cart on the other side. Further, these picking-assistant AMRs (aka “cobots”) increase productivity by pressuring employees to work faster, which can lead to employee dissatisfaction. These solutions are relatively inexpensive, very easy to install, and complement existing workforces. Because AMRs require employees to work beside them, however, they do not address the deepening labor shortage crisis facing warehouse operators.
An autonomous mobile manipulation robot, or AMMR, is a special type of AMR that includes a smart robotic arm for picking or materials transfer on a mobile platform. In contrast to cobot AMRs, these robots do not require human workers to work alongside them. AMMRs can travel autonomously into warehouse storage areas, find inventory locations, individually identify objects, autonomously pick items, and then transport them to fulfillment stations.
Advantages of AMMRs
Perhaps, the singular advantage that both AMRs and AMMRs share is their low barrier to entry for businesses who need to automate. Both types of robots are affordable, flexible, and scalable, making it easy to install automation with little financial risk. There is limited capital expenditures for this type of automation, as compared to highly expensive fixed automation systems, which usually require major infrastructure improvements.
Competing successfully in today’s retail environment means satisfying consumer demands for providing fast delivery and offering a large assortment of products. To satisfy delivery demands, many businesses install high-speed automation systems and hire additional pick, pack and ship workers. To satisfy product assortment demands, businesses continually add new products. This leads to SKU proliferation, clogging many distribution centers with highly profitable but slow-moving, low-order-quantity products. The high-speed automation and long tail SKUs are great for business — until a tipping point is reached. Businesses can quickly find themselves unable to remain competitive and unable to grow further because of two major constraints: manual labor and automation capacity.
The labor shortage is growing to be an almost insurmountable barrier as baby boomers retire in droves and the tech-savvy younger generations have no interest in manual labor jobs in warehouses. SKU proliferation has led to warehouses bulging with inventory, choking high-speed automation systems that were not designed to handle slow moving, one-item-per-order products efficiently.
Like two walls of a trash compactor slowly coming together, compressing and crushing, these two constraints — labor and capacity — are putting tremendous pressure on e-commerce enterprises of all sizes in all industries.
Here is where the advantages of AMMRs really stand out because these robotic systems act as a release valve to this pressure. How?
- Labor: AMMRs can pick and move products like humans, which means businesses can continue to add high-margin, slow-moving products without adding hard-to-find labor.
- Capacity: By moving low volume, low-order-quantity items out of high-speed automation systems and into AMMR picking modules, warehouse operators can significantly improve the capacity of their existing automation system.
Simply put, AMMRs pick more of the low-volume long tail inventory so warehouse workers and existing automation systems can be more productive.
AMMRs offer other advantages, too:
- They make better use of existing employees, further reducing labor costs. One robot manager supervising a fleet of AMMRs can handle the picking assignments of multiple workers in a manual system.
- They can boost employee satisfaction. AMMRs take on unpopular, monotonous work and eliminate the average 10+ miles a day a picker has to walk. By reassigning pickers to more engaging roles, employers can see employee satisfaction rise and turnover mitigated.
- They reduce picking errors. AMMRs are equipped with highly advanced computer vision technology, which enables the robots to see and locate objects in 3-D and in real time, providing a nearly perfect order success rate.
Investment Required to Implement AMMRs
A common benefit with all AMRs, including AMMRs, is their low cost, flexibility, and scalability, compared to other automation solutions. These robots do not require businesses to invest heavy CAPEX to buy large, fixed automation systems or to change existing warehouse infrastructure to augment their persons-to-goods fulfillment model. AMMRs, in particular, can be flexibly configured for today or the future to optimize pick, pack, and ship processes for e-commerce fulfillment. Moreover, AMMRs are scalable: Businesses can start with a small fleet of robots and easily add more as their order volume grows.
As the need for more efficient warehouses and distribution centers persist and with no end in sight to the labor market squeeze, businesses that want to compete effectively should factor autonomous mobile manipulation robots into their game plans. The firms that explore and iterate with the right type of autonomous mobile robots for their business will be best positioned to generate new levels of revenue and profit as they see their labor savings soar—all while pushing employee morale and customer satisfaction higher.
IAM Robotics created the world’s first AMMR. Retail companies now use these robots successfully to pick and transport products for order fulfillment.