Sensors remain a crucial source of data in the Industry 4.0 context. Sensors were used on manufacturing floors to collect data before the arrival of Industry 4.0 hype. Sensors can be pretty rudimental (like legacy switches and variable resistors) or advanced (like LIDAR and thermal imaging), but from an Edge AI perspective, what is critical is that they are a source of data for our Edge AI devices.
This three-part article will overview prominent types of sensors leveraged in the Industry 4.0 context in logistics. While you may see many examples of Industry 4.0 in a manufacturing context, examples from warehouse operations are scarce. So, this article will explore the types of sensors in an intelligent warehouse context.
The first part of this three-part article can be found here. In the second part, we will explore visual and force sensors in an intelligent warehouse context.
Visual sensors
Visual sensors help Edge AI to gain a sense of the surroundings or the scene around them. The key features of an image sensor are:
- Color channels
- Spectral response
- Pixel size
- Sensor resolution
- Frame rate
ACME leverages various visual sensors throughout its warehouse, ranging from tiny, low-powered cameras to high-quality, megapixel cameras. Images captured from these sensors are leveraged by AI as an array of pixel values.
On the docks, each door has “event cameras”. These cameras remain passive but act as soon as a door opens for loading and unloading. For loads that get built using automated belts, these cameras capture the count and dimensions of each package. In tandem with the barcode data on the packages/boxes/pallets, this data can provide insights into metrics like load utilization, units per man hour, loading and unloading rate, door utilization, etc.
Cameras are placed strategically on belts to identify damaged packages that move on these belts. Bot arms near packaging areas use this real-time data to place yellow “damaged” labels on these packages. The packaging bot then skips the packages with these labels, again leveraging a camera integrated into the bot.
Cameras for navigation aids are fitted across the entire warehouse to help forklift operators and pickers navigate safely and efficiently. Paired with mobile displays on the vehicles, these cameras exchange data in real-time with an algorithm that can help avoid collisions and congestion.
In another part of the warehouse, a small fleet of self-driving bots that move standardized small racks leverages a more advanced form of visual sensors known as LIDAR. These are larger and more complex (hence more expensive) “time-of-flight”[1] sensors. Though the Edge AI algorithms leveraged by ACME rarely use the data from these sensors because of the volume of the data they generate, these sensors are still very useful for these self-driving bots. These sensors provide these bots the capability to image their environment in 3D, primarily by emitting light and then measuring how long the light takes to bounce back ([1] Hence known as “time-of-flight” sensors).
Force sensors
Force sensors help measure the physicality of the environment. They are leveraged widely to capture user interaction, capture flow characteristics of liquids and gases, and measure physical pressure. Let us explore some types of these sensors being used by ACME:
Switches and push buttons: These switches often work with other sensors. As standalone sensors, they use simple buttons for human interaction. ACME warehouse vehicles have switches that operators can press to alert other vehicles when the smart camera shares a warning of an approaching vehicle in a blind spot.
Touch sensors: Smartphones and other modern touchscreens use this type of sensor. ACME captures the arrival and departure of workers through fingerprint sensing on attendance tablets in their warehouse.
Strain, flex, and load sensors: These sensors can measure the deformation of objects and the physical loads applied to them. These sensors are fitted on racks throughout ACME’s warehouse to keep track of the loading thresholds of these racks. Forklifts use load sensors to ensure safe and incidence-free loading and putaway.
Flow and pressure sensors: Flow sensors are designed to measure the flow rate of liquids and gases, whereas pressure sensors are leveraged to measure the pressure of a gas or liquid in a system (like a pipeline). ACME warehouses are intelligent buildings. The pipelines for systems like heating and cooling leverage these flow and pressure sensors for safety and efficiency.
In the third and final part of this three-part article, we will explore how ACME uses the following sensors in its smart warehouse:
- Optical
- Electromagnetic
- Radiation
- Environmental (includes biological and chemical)
The third and final part of this article series can be found here.
References:
Designed Analytics Blog 