Autonomous Mobile Robots: What do I need to know to design one?
Autonomous Mobile Robots (AMRs) augment human effort with machinery. Being autonomous, these robots perform useful work without continuous human intervention and can easily move themselves from one place to another to perform work. It has been estimated that by 2025, as many as 4 million commercial robots could be in operation in 50,000 warehouses worldwide. They will perform work like receiving, picking, sorting, packaging, and simply moving goods autonomously around a warehouse from point A to point B.
In this blog, we take a closer look at the main elements that give AMRs their functionality.
Primary Functions
A truly autonomous mobile robot must be performing at least four primary functions:
- Sensing
- Thinking
- Acting
- Communicating
Developers implement these functions using the latest sensors, algorithms, computer vision, and artificial intelligence, as well as heterogeneous computing to run it all smoothly.
The AMR below, commonly used in e-commerce warehouse fulfillment, provides one such example. This AMR integrates a range of technologies including compute, sensors, and wireless connectivity, into a form factor capable of moving large packages around.
It’s best to approach AMR design in terms of the tasks an AMR needs to execute and how it will use its hardware features and heterogeneous computing to execute them. This enables you to divide and conquer when developing the most important AMR tasks (listed below).
Sensing – Seeing Surroundings in 3D
The AMR uses sensors and cameras to not only perceive nearby objects but also understand the physical relationships among them. That means handling feeds from components like these:
- Structured light camera — Decodes a projected pattern of pixels on the scene.
- Time-of-flight camera — Measures the distance that light travels.
- Stereo cameras — Capture multiple pictures from different cameras.
- LIDAR — Illuminates a target with a laser and analyzes the reflected light.
- SONAR — Emits pulses of sound and listens for echoes.
As the AMR moves around, it uses simultaneous localization and mapping (SLAM) from camera feeds to construct a 3D map of its environment and determine its location on the map – a process known as localization.
There are two approaches to SLAM:
- Visual SLAM — Uses a camera paired with an inertial measurement unit (IMU).
- LIDAR SLAM — Uses a laser sensor paired with IMU; more accurate in one dimension but tends to be more computationally expensive.
The AMR combines motion data from the camera feeds with inertial data from sensors and the wheel encoder to better estimate motion and improve the accuracy of localization.
Thinking – Recognizing Objects and Avoiding Obstacles
AMRs have to recognize and interact with objects and get around obstacles. That means they must rely heavily on computer vision and artificial intelligence as they constantly learn to recognize objects. High-performing AMRs execute those functions on the device, instead of shuttling data to the cloud and back.
Acting – Navigating the Environment
Once the AMR has a map and knows where it is on that map, it can navigate in its environment. Navigation involves:
- Scene understanding — Using depth sensors and machine learning to build a spatial and semantic model of the environment.
- Path planning — Finding the optimal path through the environment and satisfying high-level goals while avoiding obstacles.
- Real-time control — Implementing a motion plan by translating desired speed and direction into motor commands.
- Motion estimation — Estimating a change in position on the map. With a new location and environment, the AMR updates the planned path.
Navigation includes adapting to changes in environmental elements like peoples’ movements and new shelf arrangements. AMRs rely on LIDAR to detect changes and use machine learning to refine navigation goals. They can also take advantage of Indoor Precise Positioning, using 5G Transmission Points/Reception Points (TRP) to plot a grid they can use for centimeter-level accuracy on X-, Y- and Z-axes.
Communicating – Concerting the Effort
Wireless communications such as private 5G networks in warehouses and fulfillment centers can augment a robot’s intelligence at the edge. For example, robots can leverage hybrid AI to run inference on smaller ML models (e.g., for SLAM) at the edge, while concurrently sending data to large models on edge-cloud servers for business intelligence.
Next steps
For more inspiration, take a look at some of Qualcomm Technologies’ robotics example projects. Then, head over to the Qualcomm Robotics RB3 Gen 2 Development Kit page to learn how to get started with your next robotics project.
