‘Big’ Data Game Changer

Bill Powell

Bill Powell

Department Head of Information Management, ARI

What are the trends that you are see­ing in the indus­try with respect to technology?

Rel­a­tive to the com­plex­ity of fleets, what we are see­ing most is the vol­ume of data growth, or big data as they are call­ing it these days. As fleets become more com­plex, each sig­nif­i­cant com­po­nent of that vehi­cle is pro­duc­ing data that you have to deal with. Telem­at­ics devices pro­vide pings on var­i­ous vehi­cle com­po­nents every two min­utes. It is a huge vol­ume of data.  The ques­tion is – how do you han­dle that? Not only how do you just bring in that big data, but also how do you turn it into infor­ma­tion we can use to drive effi­cien­cies for our customers?

How do you drive effi­cien­cies for your cus­tomers? What are some of the things that ARI is doing?

Right now, ARI is imple­ment­ing cut­ting edge solu­tions for cor­re­lat­ing infor­ma­tion. Within the ARI infor­ma­tion man­age­ment depart­ment is a group ded­i­cated to data inte­gra­tion. Such a team is a dif­fer­en­tia­tor in the fleet man­age­ment indus­try.  The group works directly with our cus­tomers, their ven­dors and some of our other busi­ness part­ners to cor­re­late data. As an exam­ple, for a cus­tomer with vehi­cles in thirty or forty dif­fer­ent coun­tries through­out the world, we’ll cor­re­late all that infor­ma­tion for them to pro­duce a true, holis­tic view of their fleet foot­print in an intu­itive, instan­ta­neous for­mat.  Part­nered with SAP, we devel­oped this solu­tion using unique tech­nol­ogy called HANA, which pro­vides an in-memory data­base and is a game changer not only in the fleet indus­try but worldwide.

Tell us about this.

What dif­fer­en­ti­ates this data­base from exist­ing tech­nol­ogy is its speed. It allows us to han­dle large vol­umes of infor­ma­tion as it comes in. In the past as data came in, we’d process it and then work with cus­tomers and our inter­nal per­sonal to under­stand the value of that infor­ma­tion. There were lim­i­ta­tions with how we turned our find­ings back over to our cus­tomers.  Then as the com­plex­ity of fleets con­tin­ued to evolve and the vol­umes of data con­tin­ued to grow, the return time got slower and slower. But now SAP’s HANA tech­nol­ogy has changed that. We are now equipped to not only deal with these huge vol­umes of infor­ma­tion to sus­tain growth, but we can now present that infor­ma­tion to our cus­tomers in a very effec­tive and easy format.

What type of infor­ma­tion are we talk­ing about?

The vari­ety of infor­ma­tion we han­dle cov­ers the full vehi­cle life­cy­cle from acqui­si­tion through dis­po­si­tion, includ­ing vehi­cle main­te­nance, fuel, telem­at­ics… as well as dri­ver infor­ma­tion rel­a­tive to mileage, safety and behavior.

How exactly does this dif­fer from the way you were doing it before? What is the ben­e­fit to the client?

One major cus­tomer ben­e­fit of this tech­nol­ogy is the agility that we can pro­vide.  For instance, if a cus­tomer wanted an analy­sis on its fuel infor­ma­tion, we would cre­ate a report that gen­er­ally met the client’s needs at that moment. Then, if the report needed tweaks or changes, we would repeat the process. The analy­sis itself may have taken a minute to run, or it may have taken ten hours, depend­ing on the com­plex­ity and the depth of that data. The new in-memory tech­nol­ogy can pro­duce in a frac­tion of a sec­ond what used to take ten hours. Typ­i­cally with exploratory analy­sis, one ques­tion leads to another, which leads to another. In-memory tech­nol­ogy pro­vides our cus­tomer with the agility to ask fur­ther ques­tions and get quick answers. If you’re not using sys­tems that func­tion in this man­ner, it can be a very painful process.

Can you give us a real world exam­ple of where this tech­nol­ogy has made a dif­fer­ence for a client?

In-memory tech­nol­ogy is the back-end – the com­po­nents that process the out­put.  The front-end sys­tem that our cus­tomers and our inter­nal teams work with is another SAP prod­uct that has been in pro­duc­tion for quite some time at ARI called Explorer.  Another cus­tomer story starts with us work­ing with them to under­stand their needs and require­ments for a cam­paign to reduce unpaid vio­la­tions.  We knew from expe­ri­ence that this cus­tomer pre­ferred to take a hands-on approach and be able to work with its fleet data directly.  To avoid the back-and-forth tri­als in order to get the para­me­ters just right, we empow­ered this cus­tomer by intro­duc­ing them to the Explorer tool, which allowed them to instan­ta­neously and intu­itively look at their expenses and ana­lyze var­i­ous pock­ets of infor­ma­tion.  We were able to help them iden­tify the anom­alies that were truly dri­ving those vio­la­tions and then come up with an effec­tive plan to reduce those.

Amaz­ing. Give another exam­ple besides vio­la­tions because we’re really inter­ested in this.

A few months ago, we were asked to per­form a risk analy­sis for our own com­pany to iden­tify trends and deter­mine the impact on our busi­ness.  When the request came in around 4:00 on a Fri­day, we first used the exist­ing tech­nol­ogy to per­form that analy­sis.  Roughly 24 hours later, the report timed-out.  Need­ing a more aggres­sive solu­tion, we moved the infor­ma­tion to our new HANA plat­form and redid the analy­sis. In just three sec­onds, we had the data we needed.  So as a com­par­i­son, the pro­cess­ing time went from infinity/unknown to a few sec­onds. Now that is speed!  And the byprod­uct of that speed is agility. Before the CFO was fin­ished ask­ing for the next level of analy­sis, we already had the new answer!

What does this do for ARI and sup­port­ing its customers?

We have a very aggres­sive roadmap for 2012 using this tech­nol­ogy. The first phase was to get the HANA in-memory data­base in the back-end. Next, we put in the front-end Explorer appli­ca­tion.  Now we’ve begun walk­ing through all other appli­ca­tions to fur­ther ingrain a per­va­sive busi­ness intel­li­gence plan.  An exam­ple is our Intel­lifleet® appli­ca­tion, which is in the process of being re-written.  We use this appli­ca­tion in our call cen­ter to man­age our cus­tomers’ main­te­nance and repair pur­chase orders.  A big com­po­nent of this is apply­ing the in-memory tech­nol­ogy to that appli­ca­tion so our ser­vice tech­ni­cians are bet­ter armed to pro­duce client sav­ings.  This new tech­nol­ogy will drive ana­lyt­ics into the call cen­ter – to the peo­ple that are man­ag­ing our cus­tomer spend with­out extend­ing call times.

Also on this aggres­sive roadmap for 2012 is our Strate­gic Per­for­mance Reviews (SPRs).  An SPR is a com­pre­hen­sive analy­sis that cov­ers nearly the entire breadth of infor­ma­tion that we man­age for each cus­tomer.  With this next project, we are going to extend in-memory tech­nol­ogy to ARI’s client rela­tions and sales teams so that they can make the best rec­om­men­da­tions pos­si­ble for our cus­tomers.  Cur­rently the process takes from seven to ten hours to run one analy­sis, depend­ing on the com­plex­ity and vol­ume of the data and the type of analy­sis we are per­form­ing.  Our goal is to bring this pro­cess­ing time down to less than three seconds.

After we have his­tor­i­cal infor­ma­tion and report­ing cov­ered, we’ll being devel­op­ing tools for proac­tive or pre­dic­tive analy­sis. Up to this point every­one has worked with his­tor­i­cal infor­ma­tion.  Now that ARI is work­ing at the speed of light, or so it seems, we are in an advan­ta­geous posi­tion to start dis­cov­er­ing oppor­tu­ni­ties in the very par­tic­u­lar, gran­u­lar infor­ma­tion that we are see­ing. To truly gain con­trol and drive down costs, you need to not expe­ri­ence that expense. Pre­dic­tive ana­lyt­ics can do that.  Not only has ARI invested in this tech­nol­ogy, we’ve brought on board math­e­mati­cians and sta­tis­ti­cians to build deci­sion trees and iden­tify per­mu­ta­tions.  Soon we will be in a posi­tion to tell a cus­tomer “unless you change or take a par­tic­u­lar action here, you are going to have these costs there.”  In the past, this was just sim­ply impos­si­ble to do.

One other thing that we have focused on is deliv­ery meth­ods.  Accord­ing to Gart­ner, by 2013 a third of all BI func­tion­al­ity will be per­formed on some sort of hand­held device. ARI is already there. The tech­nol­ogy we put in place in the last quar­ter enables every­thing that I have been talk­ing about to work every bit as well if not bet­ter on an iPad or an iPhone. You can have that instan­ta­neous, intu­itive access any­where you need it.

So, a fleet man­ager can be sit­ting in a meet­ing with their iPad and need some­thing imme­di­ately and have it immediately.

In the past, when a fleet man­ager needed fleet data imme­di­ately, he/she would access the vendor’s web­site and hope that the piece of infor­ma­tion is already out there. If not, then the process starts.  He/she would have to request a report, per­form analy­sis and wait for the results. That is not mov­ing at the pace of busi­ness; that is mov­ing at the pace of IT.  Now, ARI’s tech­nol­ogy allows us to empower the fleet man­ager on an iPad or iPhone to gather the infor­ma­tion, quickly change it, and visu­al­ize it until they get exactly what they need.

BIO

Bill Pow­ell, Depart­ment Head, Infor­ma­tion Man­age­ment, is respon­si­ble for set­ting the direc­tion and exe­cu­tion of ARI’s infor­ma­tion man­age­ment roadmap.  Bill started at ARI in 1991 in the oper­a­tional seg­ment of the busi­ness and tran­si­tioned to the IT team in 1995 where he was respon­si­ble for expand­ing ARI’s busi­ness intel­li­gence solu­tions.  In his cur­rent posi­tion, Bill over­sees ARI’s data­base plat­forms, its data inte­gra­tion offer­ings, and ARI’s busi­ness intel­li­gence solu­tions.
bpowell@arifleet.com

Tags: