During the holidays break last year, I got a chance to read a White paper focussed exclusively on robotics and automation, that included interviews from hundreds of Distribution center (DC) operations managers managing automated warehouses. The survey was conducted anonymously and the managers expressed their candid opinion. What was surprising was the finding that ~ 54% of those managers thought that automation has not made their job easier. 18% of them actually thought that it had added additional issues to their bucket of issues list that they have to resolve on a day to day basis.
Where is the disconnect ? On one hand, automation of warehouses is touted as a solution to all the operational challenges that a conventional DC faces on a day to day basis. On the other hand, the actual people managing these warehouses don’t think things are going that way.
In my mind, the driver behind this disconnect is not realizing the true value of an automated DC, by providing a finishing touch and converting it into a smart DC. In this post, we will explore the difference between an automated and a smart DC. In subsequent posts, we will explore some key aspects of an smart DC and how it helps support your operations strategy.
Before we embark on that journey, let us review some fundamental building blocks of an automated DC, that are also foundation of smart DCs.
Basic building blocks of an automated DC – Industrial Processes
An industrial process can be defined as a set of operations that transform, with a predetermined objective, the properties of one or more material, type of energy, or information. The industrial process also includes the collection, elaboration, and sharing of information along all production phases and steps. A very simple representation of an Industrial process is illustrated below:
Thinking about industrial processes from an automated DC perspective, a robot station that palletizes cases in an automated grocery DC is an Industrial process. The cases are the inputs, it consumes electricity and a human interfaces with its controls. The output is a ready to ship pallet.
Note that another important type of Input that an Industrial process (palletizing robot in this case) receives is Information. This information can be in various forms, like electric voltage, fluid pressure or any other type of information that can be coded in sequences of binary values.
Managing the Information aspect- Control Systems, Sensors and Actuators
Let us go back to our example of palletization robots again. Typically, Robots in automated DCs are controlled by Control Systems. According to International Electrotechnical Commission standard definition, a control and measurement system of an industrial process can be described as a set of interconnected devices communicating with each other by means of one or more communication networks.
To simplify this definition- A control system allows you to exchange information with an Industrial Process. The information exchange is both ways- A Control system sends information to the process and also receives the information.
If you think about this in simple terms, an “On” button on your Washer dashboard is part of a control system. When you push that button, it send information to the Washer that initiates the washing cycle.
What is missing in the illustration above is the interface between the control system and process. It is a device that helps propagate information from control systems to the process.
Actuators: Sending the information to the process
The actuator is a piece of hardware that transforms a command signal into a physical action on the process. It receives a signal as an input in the physical domain of the control device and sends energy as an output in the physical domain of the command variable.
But as you can imagine, this one way exchange of information, from control systems to process, via actuators, is not sufficient. Inputs to the process needs to be calibrated based on how the process is behaving given a certain input and a two way exchange is essential for that calibration. This is where sensors come into play.
Sensors: Receiving the information from the process
The sensor is a piece of hardware that is similar to an actuator, the flow of information is in a different direction. It transforms the information generated by the process into binary values, which it then feeds into the control system. These sensors are generally attached to various equipments in and around the processes (On robots, on and around conveyor belts etc.) to capture process characteristics and relay the information back to the control system.
Note that sensors and actuators are often also referred to as transducers. Since the sensor and the transducer are often physically within the same component, the two terms are often used as synonyms but best practices dictate that their usage should be differentiated by terminology.
For a more comprehensive and slightly more technical architecture of smart warehouses, please refer to my this blog post: The IIoT Data flow in a Distribution Center
The process described so far is automated- but is it smart ?
An automated DC scenario
To simplify the example, let us assume that this is a DC with only one palletizing robot, and three outbound docks. At a high level, the automated pick, pack and load process in this DC is shown below:
The process is entirely automated. Robots pick the orders and put them on conveyor belts. These belts carry the cases to palletization robot. After a pallet is created, the robot places the pallet on a belt that carries the pallet as close to the loading dock as possible. Due to the size of the pallets, the loading is then done by fork lift operators.
The Bottleneck Scenario
Note that the loading process is manual, where the forklift operators pick up the pallets from loading staging areas and load it on the trailers at the docks. Now let us assume a scenario where one of the operators gets into a minor accident which puts one fork lift out of service for a while.
If you are familiar with how automated DCs operate, you know that synchronization is the key. In this case, where loading process has now become a bottleneck, in order to avoid overflowing of outbound staging area, an operator generally will shut off the palletization process till the flow stabilizes and syncs again.
A Smart DC in the bottleneck scenario
Now let us see how a smart DC would handle this. Sensors at the loading docks would monitor the loading rate, the make sure that the flow is balanced. As soon as the sensors detect that the loading has stopped at one dock, the operating rate/throughput of palletization robot will be adjusted. The basis of this adjustment will be calculations done by an algorithm. The output value then gets to the robot via control system and actuator. No human intervention should be needed for a short term, minor disruption like this in a smart DC.
The algorithms will generally interface with a Supervisory Control and Data Acquisition (SCADA) system. SCADA systems are centralized systems that generally control and monitor the processes.
Real world is much more complicated
Unfortunately, the real world is much more complicated and DC processes are very intricate vs the simple example below. That however, makes the case for smart DCs stronger. In my next post, I will get into some details of how building the “smart factor” in your DC can help you attain the true ROI of your automation investment. We will also get into details of how a smart DC can operate with minimum human intervention, what are the aspects that we should allow algorithms to control vs aspects that humans should monitor.