PDA

View Full Version : Force sensor project with fee



Hilton
07-27-2007, 10:45 AM
Hi :

I would like a system to read the individual weight of each of 50 bottles on a shelf, on request, roughly every hour.

The system would involve the 50 sensors, fitting the bottom of the bottle to the sensor, Phidget, and the programming to download the results in a .CSV or .TXT format file to a PC.

The fee involved would make your time worthwhile.

Semicton
07-28-2007, 01:08 AM
Let's talk some more about your project :)

jdolecki
11-05-2007, 06:34 PM
Did you ever get your project done?

Hilton
11-07-2007, 12:28 PM
Hi :

Other issues came up, so the project got put on pause, but the question is still open if you are interested.

Thanks

jdolecki
11-07-2007, 04:18 PM
Are you trying to record the ammount of liquid used from each bottle?

Is there already a system/ comercial unit out on the market to compare to?

john

Hilton
11-07-2007, 04:39 PM
Hi :
Yes, the idea is to audit the bartender.
There are systems, but they work one of two ways.
1. Flowmeter in the cap of the bottle - the problem is that the flow is restricted so that it takes longer to pour, and, debris, bugs, etc., tend to accumulate - the sugar part of liquor can get quite gummy. Also, these flowmeters sometimes end up in the garbage.
2. Weigh scales, or visual inspection, done at end of shift. These are time consumimg. And, if a bottle goes from the stockroom to the serving area without being recorded, well, GIGO.

The force sensor is the major part of the project, but, in addition, there has to be some way to capture the identity of the bottle. Some bottles of liquor are $50, but there are a lot that are $10, so this is part of the challenge.

Does that answer your question ?

Regards

Dave
11-07-2007, 04:43 PM
If you're looking for an accurate force sensor for weighing those bottles, check out the FlexiForce piezoresistive sensors. With a simple drive circuit, you can get a nice linear weight/voltage response.

Top of this page:
http://www.trossenrobotics.com/store/c/2670-Force-Sensors-Pressure-Touch-Sensors.aspx

jdolecki
11-08-2007, 07:48 PM
Bottle Id I think would be the hard part. The only thing the have in common now is the UPC code.

You could use a rf tag for each bottle but then you have to apply and remove something from the bottle like the flow tops but rf tags would be cheaper. Unless they already have that type of tag.

Weighing them would be a little tricker. But an easy system would depend on the space were the bottles are stored.

What kind of money are you looking to spend for this system? And how accurate does it have to be?
Can the setup include a premade storage rack for the bottles?

john
[email protected]

Hilton
11-09-2007, 10:29 AM
Hi :

Excellent points.

A central issue in this industry is making sure the liquor is only used for paying clients.
So, the system would have to strike a balance between "absolutely unbeatable" and cost/convenience.

The RF ID tag is one solution, drawbacks being..
1. Cost $1 -$2
2. Making sure the right one goes on the right bottle
3. Telling the system which one is on which bottle
4. Retrieving them when the bottle is empty

Using the barcode on the bottle is another solution, but then the bottle would have to go back on the rack in aligned position against the reader, and there is no standard for where the bar code goes. I guess we could put on our own lables, by scanning the barcode on the bottle manually, and then produce and attach our own label.

As for a rack, when things get busy, even the most competent bartender would have a problem always putting a bottle back in the same place. And, for less competent ones, there is no way of assuring that the correvt bottle is in place.

How about the Walmart approach. Tell the supplier to affix RF ID tags OR ELSE ! LOL !

Ballpark, a 1-bartender bar can go through $200,000 - $500,000 of liqour per year, so depending on how much is not being paid for, and well you can get a complying bartender, figure 1% - 10% disappears.

On the incentive side, bars usually have different qualities of the same liquor, and the barkeeper knows this. For a larger tip, they'll pour the best stuff, or over pour. So with a better system, we can give the bartender a nice bonus based on real figures. And lots of bars have more than one barkeeper serving at the same time. A good system would keep track of this.

Regards

jdolecki
11-09-2007, 04:35 PM
Are you offering "seed" money for development , or do wou want a finished working system?

Hilton
11-09-2007, 04:46 PM
Hi :

Both. If the "seed" information worked out, we would go on to a pilot, one bottle project, then , if successful, potentailly a 20 or 30 bottle system with software to tie-in with the POS system.

Regards

jdolecki
11-17-2007, 10:43 AM
Are you a bar owner?

kdwyer
11-17-2007, 12:53 PM
Sorry to rain on your parade, but wouldn't simply hiring better bartenders be the best solution? A good barkeep not only 'overpours', but also will spring for a free drink every now and then for good customers.
While there may be 'workable' solutions, perhaps these 'solutions' introduce more trouble and expense than they are worth. ANd you sure won't make any friends out of the staff that way. Nobody likes being suspected BEFORE anything wrong has occurred.
Bartending is inherently a social task, and trying to put it on a technical measured basis de-humanizes it. People don't drink just to simply slam 'em down and get hammered as quickly as possible, and the best bars have a social atmosphere that isn't conducive to a technical analysis.
In short, some social engineering will work better than technical engineering. A good barkeep is worth his/her weight in gold.
Then again, you could always go with pre-mixed prepackaged drinks, and simply count 'em up at the end of a shift. And also lose the best customers.