Mo­ti­va­tion and de­sign

Maps are a cen­tral part of the way ge­ol­o­gists com­mu­ni­cate their un­der­stand­ing of the nat­ural world. At their best, ge­o­log­i­cal maps go be­yond record­ing the type of rocks at a spe­cific lo­ca­tion to ex­press hy­pothe­ses about the or­ga­ni­za­tion of rocks in time and space.

Lately, work­ing ge­ol­o­gists have been faced with an un­ap­pe­tiz­ing choice: to stick with a time-con­sum­ing ana­log process even in the face of ac­cel­er­at­ing map­ping de­mands and the need for dig­i­tal out­puts, or to fully com­mit to overly pro­scrip­tive dig­i­tal processes which de­grade map­ping from an in­ter­pre­tive to a tech­ni­cal task.

Map­board GIS is an iPad map­ping sys­tem for draw­ing geospa­tial in­for­ma­tion into a spa­tial data­base. It works both in a stand­alone, field-ready mode and as a teth­ered client, with a paired desk­top or lap­top han­dling geospa­tial op­er­a­tions. It is de­signed to sup­port in­tu­itive edit­ing and to make dig­i­tally com­pos­ing ge­o­logic maps a quicker, more hu­man-ori­ented process.

Pen-based map­ping

Draw­ing is an ex­cel­lent ge­o­log­i­cal map­ping method be­cause sketch­ing al­lows quick cap­ture of com­plex spa­tial pat­terns. “Stream dig­i­tiz­ing” is avail­able on other GIS plat­forms, and is rea­son­ably ef­fec­tive when com­bined with ac­ces­sories such as Wa­com graph­ics tablets, but the edit­ing tools are geared to­wards pre­ci­sion rather than flu­id­ity.

Tablet-based com­put­ing en­ables new and richer ca­pa­bil­i­ties for pen-based in­put pre­vi­ously un­at­tain­able in com­put­ing. Map­board GIS fully uti­lizes the pre­ci­sion-cap­ture ca­pa­bil­i­ties of Ap­ple’s iPad plat­form to sup­port pain­less draw­ing, eras­ing, and re­draft­ing of linework. This re­duces the dig­i­tal hur­dles to mak­ing a well-de­signed map that faith­fully rep­re­sents in­ter­pre­ta­tions.

Bal­anc­ing power and sim­plic­ity

Bring­ing in com­put­ers can add a lot of com­plex­ity to field work­flows, and care must be taken not to over­com­pli­cate the process or add new con­straints. This is es­pe­cially true when in­te­grat­ing ca­pa­ble but dense GIS soft­ware.

In ad­di­tion to stream­lined sty­lus tools, tablet de­vices (par­tic­u­larly the iPad plat­form) of­fer a stripped-down user in­ter­face that lim­its user-fac­ing com­plex­ity. iPads, much like their pre­de­ces­sor smart­phones, run in a sin­gle-app mode with only lim­ited mul­ti­task­ing ca­pa­bil­i­ties.

While open­ness and con­fig­ura­bil­ity are of­ten prized com­put­ing fea­tures, the sim­plic­ity of a tablet plat­form is a ma­jor ben­e­fit when work­ing in the field — a stripped-down dig­i­tal process is less likely to in­trude on the real-world task at hand. The user in­ter­face and edit­ing tools of Map­board GIS are sim­ple enough not to in­trude on users’ field processes.

An­other po­ten­tial side ben­e­fit pur­pose­fully sim­ple soft­ware is eas­ier use in ed­u­ca­tional set­tings. We plan to ex­plore this as­pect fur­ther in the fu­ture.

Topol­ogy-aware

One of the ma­jor hur­dles to build­ing a fully de­fined ge­o­logic map is topol­ogy, the part of the map­ping process where bound­aries are joined to fill space. In tra­di­tional GIS work­flows, this process hap­pens af­ter all linework is added to dig­i­tized and fi­nal­ized. Even with the quicker draw­ing tools pro­vided by Map­board GIS, this is a ma­jor tech­ni­cal ob­sta­cle, as it pre­vents see­ing the out­put of your map­ping un­til you have fin­ished en­ter­ing the data. Build­ing a process where changes are re­flected in topol­ogy as soon as they are en­tered, will make it much sim­pler to see the ef­fects of changes in real time.

Map­board GIS sup­ports an in­no­v­a­tive it­er­a­tive map­ping work­flow. Its topo­log­i­cal en­gine can solve the bound­aries of map poly­gons in real time, show­ing the out­put of the map­ping process as it is cre­ated. This also re­duces the risk of prop­a­gat­ing ma­jor topo­log­i­cal er­rors and elim­i­nates the “data clean­ing” that has to be done be­fore linework is closed and con­verted to poly­gons.

Field and teth­ered modes

Field map­ping is an im­por­tant work­flow for ge­ol­o­gists, but a large amount of map pro­duc­tion hap­pens back in the of­fice. Both the core data-cap­ture work­flow and the topo­log­i­cal ca­pa­bil­i­ties of the Map­board GIS can be pow­ered ei­ther by an on­board GIS en­gine, or by a con­nec­tion to a spa­tial en­gine run­ning on a net­worked com­puter. Be­cause desk­top com­put­ers can run more pow­er­ful data­base soft­ware than tablets, this has ben­e­fits for speed, data se­cu­rity, and in­ter­op­er­abil­ity. Real-time up­dates can al­low edit­ing along other Map­board GIS users and desk­top GIS soft­ware.

Build­ing new in­ter­ac­tions.

Some­what weirdly for a geo­sci­en­tist, my se­nior year of col­lege, I took a class on INLS 318: Hu­man-Com­puter In­ter­ac­tion. This course made a huge im­pres­sion on me. Frus­tra­tions we have with dig­i­tal processes are of­ten just boil down to poorly de­signed in­ter­faces be­tween user and ma­chine.

Now when I see an im­per­fect dig­i­tal process, I have the urge to ex­am­ine the process and try to find a bet­ter way. Map­ping is ripe for this kind of dis­rup­tion — re­searchers’ ca­pa­bil­i­ties are be­ing lim­ited by over-pro­scrip­tive com­puter sys­tems, and much of the prob­lem is at the hu­man-com­puter in­ter­face.

Now, I’ve el­e­gantly solved my prob­lem. So we’ve reached the sec­ond phase of this ex­er­cise, where I see if this change in ap­proach can trans­form oth­ers’ work­flows as much as it has im­proved mine.

Field ge­o­log­i­cal map­ping can be a data-man­age­ment headache, and cap­tur­ing high-qual­ity geospa­tial in­for­ma­tion over a large map­ping pro­ject is tricky. Map­board GIS al­lows quick and nat­ural cap­ture of map­ping data into user-de­fined line and poly­gon types us­ing the Ap­ple Pen­cil, with a process flow com­pa­ra­ble to work­ing atop a pa­per map. Data are saved to an in­dus­try-stan­dard open-source spa­tial data­base: Spa­tialite in “field” mode or Post­GIS a on net­worked server in “teth­ered” mode. Both of these data col­lec­tions can be opened by stan­dard QGIS or Ar­cGIS desk­top GIS soft­ware (Ar­cGIS can only con­nect to Post­GIS in read-only mode); the Post­GIS back­end ad­di­tion­ally sup­ports ad­vanced col­lab­o­ra­tive map­ping.

Cus­tom tile lay­ers in­clud­ing MBTiles files can be loaded for field-ac­ces­si­ble basemaps. Mul­ti­ple pro­jects can be man­aged on the same de­vice and shared with other de­vices. Map­ping can be done in ar­bi­trary spa­tial ref­er­ence sys­tems.

Does it re­place a GIS sys­tem?

As its full name im­plies, Map­board in­ter­nally uses a so­phis­ti­cated GIS sys­tem, but it is not a fully re­al­ized sys­tem for man­ag­ing geospa­tial data.

It pri­or­i­tizes sim­ple in­put over a fully re­al­ized set of tools. It also does not have a sys­tem for man­ag­ing fea­ture at­trib­utes.

Why?

Ge­o­logic maps are beau­ti­ful.