Allocating Resource ID for the resources in Simevents

Hi,

I am building a model of automated warehouse. This model has vehicles as resource which will serve the orders. When one order is completed the resource is released and is available for the next order. Now, i am brining back all the vehicles to one location but in real time, the vehicle will be at different locations when the order is completed. For Eg: Once order is recieved the vehicle goes to the particular order tote, brings the order tote to the operator and takes back the order tote to the original location. These locations can be different for different orders.

So when the next order is generated, i want to know the position of the previous bot so that i can calculate the distance/time required for it to reach the next order location.

Is this something that i can do with simulink. Sorry if it is a noob question.

Regards,

Tashrif Yusuf

Answer 1

[Abdolkarim Mohammadi]

I was unable to run your model, but according to your post, I think you need to calculate the distance of the orders to the operator. If this is the case, you need to define the location of the order in the warehouse when the order is generated. I have attatched a simple model where each order has three attributes: (a) order's location in the warehouse, (b) the distance between the operator and the robot, and (c) the order serving time which is proportional to the distance. You can calculate all of these upon entity generation or upon entity entry into the Entity Server.

In the attached model, I assumed that the operator is located at coordinates [0,0,0] and the location of the generated order in the warehouse is a random triplet [x,y,z] in the range of [1,1,1] to [5,5,5]. I calculated the distance using vecnorm function, which calculates the Euclidean norm. I also assumed that it takes 2 seconds for the robot to travel 1 in the warehouse. You can customize these assumptions to match your problem. The model also outputs the attributes, which is accessible via the variable out.SimOut in the base workspace.

[Tashrif Yusuf]

Thanks a lot for taking time to reply and building a model for me with such detail and calculations and the formulas. The reply is very much appreciated as i was finding it very difficult to get any help doing it alone as i am a begginer in Matlab/simulink and simevents.

I am not sure why you were not able to run my model, is it showing any specific errors? I tried again and it was working for me. I was caluclating the time required for the motion using the physics equations of motion in matlab and have created a function for that. So, when i specify the location in x,y,z (x is vertical, y horizontal and z is the levels) and the operator id which item has to go to it will return me the service time for that. I was successful in doing that and calculating the throughput.

The normal time consideration is done based on the bot taking the bin out from the rack and coming to the operator, operator takes the item from the bin and bot returning back to that same location. So, the problem with the model is either i have to make all the bots move to one fixed position after the process is done so that i can calculate the accurate time.

But in the real world, once the bin is replaced the next order may be the adjuscent bin, so the bot have to travel only short distance to reach there. What i assume happens in the simevents model is that once order is completed the bot information is lost. The ideal time would be the last position of the bot from the previous order and the time it takes to move from that position instead of a fixed position.

The above mentioned was one of the issues i am facing. I hope i was clear enough.

Also, i am running Matlab 2020 version, is that reason you are not able to run my model? Should i export to the file to older versions?

[Tashrif Yusuf]

Also, i have created a small PPT to give better idea of the model i am trying to build. There are multiple rack locations shown as 1,2,3,4 and multiple operators for serving the orders. The vehicle takes a bin and comes to the operator and the operator takes the item from the bin and the vehicle with the bin goes to the rack.

There is a turn table which help the vehicle move from operator to the rack location. For example, if the vehicle has to move from the operator 1 location to rack 4 location it will move to the first turn table and the turn table rotates 90 deg andit moves in the y axis (green, horizontal) and passes 2nd, 3rd and reaches 4th turn table rotates 90 deg and moves on to the rack location.

The output i am looking from the simulation is the number of bots that might be required if i have to achieve a throughput of 1500 item processing per hour.

[Abdolkarim Mohammadi]

I think I am understanding your problem. Correct me if I'm wrong. You want the bot to have access to the information about all of the pending requests and make decision whether or not to serve two adjacent orders simultaneously and bring both of them to the operator?

[Tashrif Yusuf]

Thanks for the reply. That would be definetely something i want in the future bur righit now my problem is a little different.

Consider a first order placed which is in rack 4 coming to operator 1. I'm considering the origin point as the location of the bin and considering the travel time to the operator and return time to the same point on rack 4.

Now, if a second order is generated which is at rack 3. I'm assuming the bot is available at rack 3. and calculating the time. In real world, the bot has to travel from the previous order location which is at rack 4 in the example case and move to rack 3.

The one way in which i can solve this is send all the bots to a fixed location on the system and calculate the time. But, this won't be a realistic representation of the system.

So, for calculating the distance or the time required i need to know where the previous location of the 'finished order' bot is. I can't figure out any way to do this with the limited knowledge i have on the tool.

[Tashrif Yusuf]

I found out a small workaround to this problem if the system is having a single bot. I created an array in matlab with the first coloumn the serial number, 2nd, 3rd and 4th coloumn the x,y and z location for the particular order. If i have 360 orders, i created 360 lines with the x,y,z information in an array called order_details.

This is the code i have in entity generator. If the entity id is matching with my order_details array then i give the value of x,y,z. After finishing the order, i am sending it to return path which will be 'yreturn', 'xreturn', 'zreturn' details which i can take from the next order in the queue as shown in the last line of the code.

I believe this will work for a system with only one bot because the same bot will have to come to service the next order. But, if there are more bots in the system i am not sure how to do this.

Hope i was able to explain the issue i have. Do you think there is a way in simevents to solve this isssue?

[Abdolkarim Mohammadi]

As far as I could understand your problem, I think you have multiple bots going around and serving orders, and one or multiple pending orders at the location of operators. You want all of the bots to memorize the location of previous orders. I think there are two ways to do this:

1) Attach an attribute to the bot named as PreviousOrderLocation. As soon as the bot reaches to rack 4, we have

entity.PreviousOrderLocation = 4; 
entity.CurrentOrderLocation = 3;

2) You can also use set and get functions like SetPreviousOrderLocation and GetPreviousLocation to retain the location of the previous order. When the bot reaches to rack 4, we have:

SetPreviousOrderLocation (1, 4); 

Where 1 means the first bot and 4 shows previous location for the first bot. You have to set the output signal dimension of the SetPreviousOrderLocation and input signal dimension of GetPreviousLocation to [4,1].

And when you want to calculate the distance elsewhere, you will have

PreviousLocation = GetPreviousOrderLocation (1); 

to retrieve the previous position and perform further calculations. Hope this helps.