Slashdot Mirror

← Back to Stories (view on slashdot.org)

Electronic Service Signature Solutions?

Posted by Cliff on from the digitally-obtaining-one's-john-hancock dept.

LilJC asks: "I work at a company that does service work nationwide for a major chain. Our vendor requires a signature for every visit to every store. We have been running thus far from ICR on faxes and linking to the faxed signature image on our web site, but are having numerous problems trying to depend on a thousand+ faxes a month coming through perfectly. We are looking for an economical alternative that would allow a contractor in the field to get a signature and transmit it to us in an system I can idiot-proof. Something along the lines of the UPS guy's digital clipboard, but without the pricetag of a Telxon. Brainstorming myself, I wonder if there must be a cheap PDA/modem combo or the like that would allow a contractor to enter required visit information, get a store signature, go home, plug it into their phone line (without needing a computer), and hit send. I can code and am not afraid of doing some custom work for an app on a handheld. Does the Slashdot community know of any projects along these lines, or have ideas about a wise choice for hardware to homegrow the software on?"

22 comments

  1. Can you afford it? by pbox · · Score: 1

    Can you afford to have the contractor
    a) lose un-uploaded PDA
    b) have the battery run low
    c) other catastropes...

    If you can , then it is ia good idea. Make sure to GPL it, pleeez?

    --
    Code poet, espresso fiend, starter upper.
    1. Re:Can you afford it? by pbox · · Score: 2, Interesting

      One more thing, would it be not easier to write a fax server system, in which your contractors can fax in the signatures, and it would organize it and archive it automatically?

      You can have the field contractors dial the server's fax number plus the order number, and the server would store it accordingly.

      This would utilize already technology that is present everywhere (ie fax machine) and would not put any burden onto your field contractors (in regards to carrying/learning/purchasing new equipment/software)

      --
      Code poet, espresso fiend, starter upper.
    2. Re:Can you afford it? by LilJC · · Score: 1
      This is essentially what we are doing now. The contractor's ID, the store number, etc are written on a report that is signed and faxed. Form recognition/ICR software dumps it into a postgres db and our web server generates pages off that. The system works great 90% of the time, but unfortunately it's very difficult to get that last 10% to fax in their reports right-side up, re-send them if they jam, or just send them at all.

      90% might seem good for most applications, but when you've got say a 15% margin and lose 10% of that, it's suddenly worth the risk of losing a cheap PDA now and then.

      --

      The only thing more dangerous than a file named -rf is renaming it -rf\ /
  2. FWIW.... by Anonymous Coward · · Score: 0

    I've heard that Microsoft's Tablet PCs are pretty damn good at handwriting recognition (actual handwriting, no PalmOS special shit). I know you said no new computers, but a cheap tablet pc w/ a custom app written by you might be worth it.

    or, just get any ol' pda w/ built in modem and api for creating small images with the stylus.

  3. surprisingly easy to do. by Kevin+Stevens · · Score: 1

    I did a bit of work like this on a palm pilot like device a few years ago. It was essentially a PDA w/ a laser scanner built into it made by symobl. It ran Windows CE- whose SDK is real nice, and makes it simple to make apps. the modem thing... that might be a little tricky- and expensive. How about USB? and the handwriting recognition API may need to be licensed- you would have to look at specific devices and platforms. But in essence, yes this is very doable, and probably not that hard. I was surprised how easy it was to program these things- there is nothing low-level involved at all.

  4. Some basic thoughts by coyote4til7 · · Score: 1

    This does sound like a nifty project.

    The things to evaluate your platform choice by (OS in general, specific device in particular) would seem to be cost, dependability, battery life, ease-of-development, etc., I suspect B&W Palm devices are going to hit it best with the possible exception of development. There are probably Palm-licenses that sell the hardware combo you need.

    You really want to look at getting devices with some sort of wireless internet provider. Write your custom code to every so often check for an internet connection & if found, upload the data.

    You should really look at having cradles that mount into the vehicle and ride off the car's power.

    And think through the entire process these guys go through to make this fit as simply into the way _they_ work as possible. Or you might find out you've got a lot of custom paperweights.

    Pushing the GPL would be wise. There's nothing like cutting costs and getting free publicity down the road as the app spreads and people jump on board...

    --

    the clock on the wall says 4 til 7
  5. Simple... by dJCL · · Score: 1

    From your description, you need to log what was done with a signature from the location's manager to verify it was done... simple.

    Example: with a pocketPC based system, just log the data in a excell spreadsheet and save the sig as an image...

    or

    Code something yourself, the tools are free from MS, either vb or C...

    And for the backup issues, just give them a media card that they can backup and make it mandatory that the card be removed in the program, once the data is updated, then one copy will be relativly seperated from the other if one is lost and pccard of compact flash card modems exsist if you want to communicate directly... just throw together a simple webserver to deal with that, get a few accounts with nationwide ISP's and you are set...

    Not really that complex... If you offer me money, I'll write if for you...

    --
    On Arrakis: early worm gets the bird. Magister mundi sum!
  6. A possible solution... by MrIcee · · Score: 3, Insightful
    I'll make an assumption about your problem here that may not be true. But here goes... assuming that by *signature* you mean 'a verified way of determining who went to the store'... and that perhaps other things than the signature could be used.

    One possibility would be to issue each vendor and each employee that requires tracking a unique bar-code ID. Then a simple scan tool, one carried by the employee and one at the vendor could be used. The vendor could scan the employees ID and the employee could scan both the vendor and his own ID and that info could then be easily uploaded and visualized in a large number of ways. That would give you verification in both directions, as well as a timestamp of exactly when the employee visited.

  7. Royal Linea by jayrtfm · · Score: 2, Informative

    the Royal Linea may be a good platform for you, since it has a built in modem, and is reasonably priced.

  8. Visual Studios.Net and the compact framework by Duwke · · Score: 1

    Have you looked at the compact framework? Your application would be quite easy with a .Net PDA and modem combo. It's amazing how fast you can put something together, even if you just want to prototype something.

    It would be an interesting project to try and get the application to use a (compact flash) modem such that the end-user just plugs it in. It's probably possible to automate the modem connection with the PocketPC SDK (then wrapped with com interop).

    Anyway, neat project. If you need any help, let me know!

    1. Re:Visual Studios.Net and the compact framework by Duwke · · Score: 1

      Just out of curiousity, I wrote the part where it takes the signature and writes it to file (bitmap). Less than 50 lines of code. It's almost too easy...

  9. My sig by Anonymous Coward · · Score: 0

    Whenever I encounter a signature-gizmo, I just sign my name, "Richard M. Nixon" and no one has ever noticed or said a word.

  10. Security guard tags by martin · · Score: 1

    What about those little devices that security use to prove they've visited various parts of the site? Would be fairly cheap...

  11. Pick two: by jolshefsky · · Score: 2, Insightful
    • Features
    • Cost
    • Development time

    Pick two:

    • Features: allow a contractor in the field to get a signature and transmit it to us in an system I can idiot-proof
    • Cost: economical alternative

    In other words: this may take a while.

    --
    --- Jason Olshefsky

    Karma: Poser (mostly affected by adding this line long after everyone else did)

  12. My dumb suggestion... by MythMoth · · Score: 1

    Mobile phone ?

    Presumably there's something similar to the Sony Ericsson P800 available in the US.

    That and a bit of custom software should do the trick. It'll be pricey to get someone else to write it, but if you're able, all the dev software you need is freely available.

    If some of your customer's stores are out of mobile range, fallback to fax.

    D.

    --
    --- These are not words: wierd, genious, rediculous
  13. Why go electronic? by duffbeer703 · · Score: 1

    If you cannot afford to buy the expensive field equipment you need, just use paper forms. Go to Costco or Sam's Club and by a compact fax machine and bag for $40.

    --
    Conformity is the jailer of freedom and enemy of growth. -JFK
    1. Re:Why go electronic? by cdrudge · · Score: 1
      Next time spend the 15 seconds to read just a little bit of the question. Third sentence down:
      We have been running thus far from ICR on faxes and linking to the faxed signature image on our web site, but are having numerous problems trying to depend on a thousand+ faxes a month coming through perfectly.
      They are using faxes. But keeping track of 1000+ faxes and hoping that they all come through clearly is a problem.
    2. Re:Why go electronic? by duffbeer703 · · Score: 2, Insightful

      They need to look at their forms and the fax system.

      Improving the fax backend would be infinitely cheaper and more reliable in the end than moving the problem to the field.

      --
      Conformity is the jailer of freedom and enemy of growth. -JFK
  14. Custom solutions by penguin_nipple · · Score: 1

    It's a simple solution, much simpler than one might think
    Here's the idea:
    Capture the pendown event, ie every time a styles hits a specific area of the screen
    Record the x-y co-ordinates of the event
    Record the co-ordinates to a file
    Write a simple routine to redraw the co-ordinates at a later date
    Simple, and yes I have used this technique numerous times, it is much better than the Handwriting API as it records the actual points and you can do co-ordinate doubling for a faster response to the pen down event.

  15. We're doing this... by knewter · · Score: 2, Informative

    Actually the company I work at, InLine (http://www.InLine.com) is doing this currently. I'm developing the app right now (in .net), and we're almost live with it. We also tie to a customer database, which keeps track of various assets the customer has with us (an asset-tracking, trouble ticket system, basically). Rather robust, if I might say so. Alas, no chance of open-sourcing this project - we've spent a considerable amount of time refining it.

    Anyway, point is, it can be done, and relatively easily. But it will take time to do it right.

    --
    -knewter
  16. Open Source J2ME + Palm solution by scottdavis99 · · Score: 1

    I have an Open Source solution that might meet your needs.

    We captured signatures using J2ME and off-the-shelf Palm devices. We persisted the sigs as SVG (Scalable Vector Graphics, an XML dialect -- http://www.svg.org/) in a local datastore. Later, we used a custom JSync Conduit to slurp the data off the devices and upload it to a waiting server during the HotSync process. The server (Tomcat -- http://jakarta.apache.org/tomcat) ran Apache Batik (http://xml.apache.org/batik) to convert the SVG into JPEG for web presentation.

    We used Apache HTTPClient (http://jakarta.apache.org/commons/httpclient/) and Jason Hunter's com.oreilly.servlet (http://www.servlets.com/cos/) to do a MultipartRequest file upload. This allowed us to provide a more traditional HTML form to upload the sigs as a backup if the conduit failed. You could just as easily use Apache Axis (http://ws.apache.org/axis/) to catch the upload as a web service.

    All of the technologies involved are open source, so you can string them together yourself.

    Or, you can grab some rough source code as a jumping off point. I was very clear about my intentions to open source the J2ME app from the start of the project. I named it "Scratch" (as in "what was itching me at the time...") Your post motivated me to get it online (https://gen4j.dev.java.net/servlets/ProjectDocume ntList?folderID=0). Long term, the project should be accessible at http://scratch.dev.java.net.

    Currently, I just have the client app posted. Sadly, there is no documentation at this point (but there will be soon, I promise!). It's only 5 classes, so I hope that their roles will be self evident.

    The conduit needs some serious refactoring. That will get posted w/in 30 days. I am presenting a case study/how-to on this very subject at two different conference in Sept, so this code will be at the top of my list of ToDos...

    Good luck, and hope this helps.

    --
    ~Exaggeration is a billion times worse than understatement.
  17. very last post by Ruby+last+post · · Score: 1

    Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins, Grumpy Watkins,

    --
    Ruby Sleeps