I work for a small manufacturing company. We own a few buildings and need to transport raw and completed goods by cart between stations within the buildings and across a parking lot to the other building where more stations are.
We have determined someone running parts back and forth is going to waste time. We priced out AI powered carts but they are as much as 40k each. Being a small business, there is no budget like that.
Our goal is simple:
Have the cart/s go from station to station with a control panel that has those stations pre-programmed. Building 1 could have 5 waypoints where it would stop to get parts. Same for building two. We need it to know where those points are, and how to get to the other building.
I have looked all over the internet and can't seem to find a build like we need. I see line following bots but they do not have waypoints to stop at. It looks like this tech can use obstacle avoidance sensors to avoid people and things as well but can it use LIDAR or something more advanced but also pretty affordable now days?
We would love to have 4 to 6 of these carts made and have them know where to go and such. If we did a line we would want them to know how to reverse if they were beside a station and needed to go back vs all the way around the loop.
Our carts can have boxes up to 8ft long and need a capacity of at least 500 LBS. We have a metal fab shop to build any of the metal parts needed. We also have wheelchair motors that might could work for the power. A lithium battery could be sourced. We just don't know about all the rest.
Nobody seems to have a cart like we need. A lot of the on the market solutions cost a lot of money and carry small items. I know toyota has a line based cart but its likely very costly.
We could start off with 1 cart and see how it goes. We were hoping to build it on a small budget. Does anyone feel it can be done for under $2000 in parts per cart?
Let's see it I understand this! Your company owns several buildings and leasing a system of moving self-powered carts is not in the budget? Consider how much your companies insurance will increase when you have a system you have built is operational. Bet the lease payments will be less than the insurance increase.
Have you discussed this with your company owners?
A leased system could be operational, yet, this year. How long will you take to get your design operational? And you need to design from the ground up, not from the controller down. Get all the mechanical stuff done first.
Not meaning to disparage your interest, but I have been part of several small businesses and I would never allow you to design and build such a system for my company. Too great of a risk.
If a device is receiving 4 RSSI signals at a point those RSSI signals will represent a unique position location. If the device is moved what happens to the RSSI signals? Why they change based upon the new physical location, right?
Consider the traffic control needed for this option. What if a cart is en route and another cart wants to go back on the same line? I guess it's possible but the in-floor guidepath systems I've seen have the vehicles all going in one direction - generally clockwise if viewed from above.
Well honestly that was the next question. Safety is important. Having LIDAR and obstacle avoidance sensors on the unit all around would still have some risk?
I just wanted to know if this was even possible and what the basics was. We are simply pondering some outside the box ideas for a goal.
Budgets are tight currently. Just finding workers to show up now days is hard, no matter the pay. Do you know a leased system like this that does not cost 100k to set up?
That is quite cheap by any measure. Have you actually spent time talking to the companies? They are as anxious to sell as you are to buy. Remember, the system will operate for up to 10 years. Have you considered the cost to develop your system and to maintain it operationally from day to day, 365 days a year? How many new employees will be needed to maintain and operate your system?
You make valid points. My goal with asking this is 1: If it was possible and 2: if it was a good idea.
I get the feeling it might be more work than it's worth and will have safety/liability issues a leased or owned commercial unit won't have. I do appreciate everyone's thoughts on this. Thanks so much for voicing your input and being helpful.
Sounds like a fun project. The biggest challenge is the mechanical one. Making a toy sized version for proof of concept should be trivial. I would probably opt for a burried wire for guidance but there are all sorts of options. I would also think about maybe repurposing mobility scooters. Take the seat off and put on a large bin.
I would also consider low voltage electrified rail and dedicated guidway.
You are not going to build this system for under $100k. Not unless everyone involved works for free. And even then...
OK, sure you can build a proof of concept and have some fun. But think about it. Is building AGV's something that your company wants to get into? Then there's the ongoing maintenance, the bug fixing, the "this only happens at noon on Friday" problems, etc.
In the end, is this really going to cost you less than a minimum wage employee in a golf cart?
About ten years ago I built a stage prop out of a handicap cart that had to repeat its movements. It did not use an Arduino, but it would have been much easier than what we did.
So, there's a suggestion for your basic frame. We used something like this:
The seat and steering were removed and the prop was built on the frame of the scooter.
Not a problem. Use the WiFiMulti library. If the Arduino loses WiFi connection it will look through its list of access points to find another one. You could probably stop forward motion when this happens.
RSSI levels are useless for positioning. People walking between the device and the access points will make the RSSI change.