Fog Creek Software
Discussion Board

Field Data Collection Application

Hi all,
I have to develop a mobile data collection application which will be deployed on handhelds and the data collected will be synced to a central database server via modem on the handheld.
I have done the similar stuff using Palm devices. But this time the field condition is very rugged and the filed people are not very computer savvy. I am afraid that the Palm won't be very durable in these conditions. I have looked online and all the rugged handhelds I could find are very expensive, like $1000 a piece. But if I use Palms, I could buy one for $100.
Has anybody here implemented a similar field data collection system? Any ideas or thoughts? Thanks for all your inputs.

Sunday, January 26, 2003

Hi Yaniv,

In my former engineering consulting company we used to use handhelds for field data collection. I built a number of different systems, using a number of differnet devices. From my experience, we ended up finding that standard devices worked best, and were certainly more cost effective. They're fairly tough to begin with, barring extreme moisture.

The biggest issues that we learned were 1) Use a device with a color screen. Field technicians don't like having to deal with the ugly and sometimes hard to read mono displays. 2) Field techs by far were more productive, comfortable, and happy with WinCE/PocketPC based systems. What little computing knowledge that they have is based on their Windows desktop systems. That's how computing makes sense to them, so by and large that was what they were most comfortable with.

If the devices are going to get wet, then you need to go rugged, or at least a rugged case over a normal PDA. Otherwise I'd recommend going with the smallest, lightest PocketPC device around. The smaller and less obtrusive, the better for the techs. The new Toshiba is nice, about the size of a Palm V, but it doesn't have a modem or compact flash card slot. I understand that Compaq has a new small iPaq coming out, but I don't know the specs.

The other way to go is to standardize on one of the original iPaq models and then buy them all used or NOS from eBay or the like. It's much cheaper that way, and when one goes down, it's cheap and easy to replace.

Oh yeah, make sure that you are using a unit with a reflective screen. It makes a big difference in varying light conditions in the field. Also, we found that saving data to a compact flash card during operation was better, as a protective measure. If something does happen to the unit, it's much easier to move the card to a new machine and keep going. If it's in the unit's data store, it could be much harder and result in some lost data.

Hope that helps, let us know how it turns out.



Sunday, January 26, 2003

Hi Josh,
Thank you very much for your input.
Pocket PC's are nice and easy to develop but little bit expensive though. But you mention a very nice point about the color screens though. Right now I am leaning towards cheap Palms with some rugged cases. Any pointers?
I have been using CodeWarrior for my Palm development so far. Have you used Satllite Forms? What you think of it?
Regarding adding CF card to every field unit, kind of adds up to the cost. We are planning on giving modem with every Palm anyway, which the field technicians will use to sync the data every night. What do you think of it?
Thank you once again.

Monday, January 27, 2003

Yaniv -

I've got to second JWA's points. We've implemented such an app on a handheld platform. It's WinCE, on a moderately ruggedized unit.

A couple of points: absolutely get the color screen. We've got a mono screen on ours (we lost that argument with the project sponsor), and we can see that when it gets in the hands of the end users, they're going to have trouble seeing the darned thing. Hell, we have trouble seeing it ourselves in indoor office lighting, and that's with the backlight on all the time. The unit's operating environment will rarely be outdoors; it will be primarily indoors in heavy industrial facilities, though at outdoor temperature and humidity.

yes, these ruggedized units are pretty expensive -- each handheld unit, presumably including the cradle and power adapter, is over 1000 USD each.

Something to think about too is size of keys. Granted the smaller form factors may be more convenient, but if your users will be outdoors, or like ours in often unheated though sheltered areas, they might be wearing gloves. Handling a stylus and hitting small keys could be hard with gloves on. Also, this unit is intended out of the box to be used with a stylus. However, the task the users will be doing with the units will not let them have a free hand to use a stylus, so we had to make the application completely based on key-entry and key-navigation. There was a bit of work to ensure we disabled the stylus during normal operation and remapped the keys so that the user could use the larger keys rather than the small keys for all necessary functions.

It was an interesting experience for us, definitely something new and different for me. Good luck with it.

Monday, January 27, 2003

Oh yes - and second also JWA's point about making sure dynamic data's getting stored on the memory card during operation, for reasons he mentions. All sorts of things can happen - battery dies, falls out, somebody fat-fingers the on/off switch wrong and reboots the unit, etc.

Monday, January 27, 2003

"Regarding adding CF card to every field unit, kind of adds up to the cost. We are planning on giving modem with every Palm anyway, which the field technicians will use to sync the data every night. What do you think of it?"

Jeeze - sorry Yaniv, last thing. regarding your comment cited. Personally, I think it's a mistake; I think the risk of data loss it too great--get it into something permanent like the card and into the main server ASAP after collection. To evaluate the cost of the card, what's the cost of one of the Field Tech's days collecting data? If the info isn't on a CF card, and something goes wrong with the upload -- or it isn't uploaded till end of day, then what if you have a failure at or near the end of the day and the tech loses the entire day? How expensive will that be? Obviously I can't answer for you, but stack up the cost of the cards against the cost of a tech's work for one day because that's basically what you're talking about losing.

Monday, January 27, 2003

Thank you very much for your inputs.
The users of this application will be generally outdoors and will be wearing gloves.
I do understand your points about color screen. But all they will be accessing is some data collection forms. Won't  the color screen will be an overkill? What about a backlit mono display?
And regarding them operating with gloves on, how about some rugged cases for Palms and some external thick stylus?
Also what kind of software development platforms you used in the applications you guys developed?
Once again thank you very much.

Tuesday, January 28, 2003

Well, to respond to your comments, Yaniv -

what we've got that I said we already have trouble seeing now--in the office-- IS backlit mono screens. So, if it were my decision, take it for what you want, yes, I would go for the color. I think color is important because those displays are normally brighter and you can play with different color contrasts. Even if you only use two colors, fg & bg, you can play with different color sets if you permit that to be a user selection in your app. Some folks might find black on yellow easier to see in bluish lighting than some other default color pairing, for example. The point is, you have better flexibility.

You're building this application to improve the process of data collection and entry, which means less time, better accuracy. If the tool is at all hard to use in its environment, then you're going to sacrifice time, accuracy or both and the whole thing will have been pretty pointless.

As I said, if you know your guys will be wearing gloves, then keep that in mind both in the selection of the instrument (after all, Blackberry devices are great, but how would you like to try to type an email into a Blackberry with gloves on?) the types of keys/controls, and how you might remap controls to make life easier and less error-prone for the users. Unless they have really good gloves, their hands are going to get cold, too, so expecting really good dexterity is probably a gamble.

Look, Yaniv, - I don't have a clue who you are or who you work for, and I'm certainly no expert in this field, but if you want some authoritative advice in the area of human-machine interface for situations like this (not just talking computer GUI's here), drop me an email and I can put you in contact with somebody I know quite well who maybe can give you some useful advice or point you in the right direction -- it's your choice; you're welcome to contact me via the email link below.


Tuesday, January 28, 2003

Sorry for the delay in posting. Thank you very much for your advice. I am still negotiating/working with the cliet to get this project started. Once the contract is in place, I would love to get in touch with you, if I need any help. Thank you once again.

Wednesday, February 05, 2003

No sweat, Yaniv - I'll be here. Good luck.

Thursday, February 06, 2003

*  Recent Topics

*  Fog Creek Home