Data at the heart of supply chain resilience

The severe cir­cum­stances of the Covid-19 pan­dem­ic forced organ­i­sa­tions to reflect on the process­es that are yet to reach dig­i­tal matu­ri­ty, hold­ing sup­ply chains back. The sud­den shift in con­sumer demand, cou­pled with lock­down-relat­ed pro­duc­tion out­ages, cre­at­ed a high­ly vis­i­ble sup­ply cri­sis doc­u­ment­ed by images of emp­ty shelves in shops. Less vis­i­ble impacts were felt even hard­er by busi­ness­es.

Despite their best efforts to adapt as best as they could, the lack of vis­i­bil­i­ty and flex­i­bil­i­ty with­in sup­ply chains was already so deeply entrenched that delays were almost impos­si­ble to avoid. Frag­ment­ed IT infra­struc­tures, over many years, have cre­at­ed data silos that pre­vent ready access to crit­i­cal infor­ma­tion when it mat­ters. An inabil­i­ty to make informed deci­sions quick­ly, based on the most up-to-date infor­ma­tion, hin­ders organ­i­sa­tions from respond­ing to sud­den sup­ply issues or changes in demand, leav­ing com­pa­nies scram­bling dur­ing times of dis­rup­tion.

“There has been sig­nif­i­cant sup­ply dis­rup­tion in both the first mile and last mile of demand ful­fil­ment,” says Mark Holmes, senior advi­sor for sup­ply chain at Inter­Sys­tems, a lead­ing provider of data tech­nol­o­gy. “Keep­ing man­u­fac­tur­ing up and run­ning has been incred­i­bly chal­leng­ing, as well as rebal­anc­ing inven­to­ry to meet cus­tomer demand through dra­mat­ic changes in buy­er behav­iour. The shelves have often been either bare or not the typ­i­cal prod­ucts we like to see.

“All of this has high­light­ed the inad­e­qua­cies of a lin­ear non-dig­i­tal sup­ply chain, as well as inef­fi­cien­cies in pre­dict­ing and man­ag­ing the impact of con­straints in the sup­ply chain. These inad­e­qua­cies were always there but have been exposed by the pan­dem­ic. Com­pa­nies need action­able insights in a more col­lab­o­ra­tive and dig­i­tal sup­ply chain. The only way to get out of fire­fight­ing mode is to real­ly accel­er­ate dig­i­tal trans­for­ma­tion and improve sup­ply chain vis­i­bil­i­ty.”

In a recent Inter­Sys­tems sur­vey of retail, con­sumer pack­aged goods and man­u­fac­tur­ing organ­i­sa­tions, 83% of respon­dents said the pan­dem­ic wors­ened their tech­nol­o­gy-spe­cif­ic sup­ply chain chal­lenges. The most preva­lent prob­lem is the lack of flex­i­bil­i­ty in exist­ing process­es, cit­ed by 43% of sup­ply chain pro­fes­sion­als, but many are also expe­ri­enc­ing oth­er data-relat­ed issues, includ­ing a lack of access to cur­rent data and accu­rate end-to-end vis­i­bil­i­ty, and dif­fi­cul­ties per­form­ing ana­lyt­ics and inte­grat­ing and nor­mal­is­ing dis­parate data.

All of these chal­lenges exist­ed long before the Covid-19 cri­sis and will con­tin­ue to cause prob­lems when the pan­dem­ic final­ly ends, if organ­i­sa­tions don’t improve their data man­age­ment and over­all sup­ply chain vis­i­bil­i­ty. Com­mer­cial off-the-shelf soft­ware and siloed sup­ply chain appli­ca­tions are not suf­fi­cient. An addi­tion­al data man­age­ment lay­er is need­ed to enable organ­i­sa­tions to over­come the issues they are fac­ing, remov­ing silos with a sin­gle, accu­rate and cur­rent view of data that helps them gain bet­ter vis­i­bil­i­ty inter­nal­ly and across their sup­ply chain.

“The data prob­lem is spread all over,” says Holmes. “There is no sin­gle source of truth, and it’s not a new prob­lem. Pre­vi­ous approach­es to data man­age­ment and inte­gra­tion have been chal­lenged with the com­plex­i­ty of mov­ing data in batch mode, and man­u­al­ly using that data. Dis­parate data sys­tems are hold­ing back the entire sup­ply chain ecosys­tem, which the end cus­tomer relies on to get the prod­ucts they want. Putting data at the heart of sup­ply chain resilience allows a com­pa­ny to orches­trate dis­rup­tion and con­straints of sup­ply and demand with­in the sup­ply chain to meet ful­fil­ment in a pre­dic­tive man­ner with a high degree of accu­ra­cy.

“It’s exact­ly the prob­lem that Inter­Sys­tems has been solv­ing for decades. We con­nect dis­parate data sources and enhance vis­i­bil­i­ty. It’s in our DNA. We iden­ti­fy crit­i­cal busi­ness ini­tia­tives, includ­ing under­stand­ing the impact of vari­a­tions in demand and sup­ply, opti­mis­ing replen­ish­ment cen­tres and inven­to­ry rebal­anc­ing. Our archi­tec­ture is built around a smart data fab­ric, an excit­ing new approach which allows us to con­nect to core sys­tems and an ecosys­tem of sup­ply chain part­ner sys­tems, to access data on demand while also reduc­ing com­plex­i­ty and accu­ra­cy and improv­ing the cus­tomer expe­ri­ence.”

Process­es such as demand fore­cast­ing and inven­to­ry opti­mi­sa­tion can be fur­ther enhanced with the use of the lat­est arti­fi­cial intel­li­gence (AI) and machine learn­ing (ML) tech­nolo­gies. Acquir­ing large num­bers of skilled data sci­en­tists is an unlike­ly option for most organ­i­sa­tions, least of all when such tal­ent is so dif­fi­cult and expen­sive to recruit. Busi­ness­es must there­fore con­sid­er how they can cap­i­talise on the ben­e­fits of AI and ML with­out the require­ment for cost­ly expert knowl­edge.

New inno­va­tions, such as Inter­Sys­tems Inte­grat­edML, allow organ­i­sa­tions to eas­i­ly add advanced ana­lyt­ics to appli­ca­tions with­out the need for such in-house exper­tise. The tech­nol­o­gy sim­pli­fies the process of build­ing, test­ing and deploy­ing ML mod­els, and auto­mates the process of inte­grat­ing them seam­less­ly into pro­duc­tion appli­ca­tions. Inter­Sys­tems Inte­grat­edML helps busi­ness­es to devel­op accu­rate ML algo­rithms direct­ly with­in exist­ing data man­age­ment infra­struc­ture, with­out requir­ing a team of skilled data sci­en­tists, while embed­ding the mod­els into sup­ply chain appli­ca­tions to take pro­gram­mat­ic actions in response to real time events.

“Intel­li­gent data solu­tions, and tech­nolo­gies like AI and ML, will pow­er sup­ply chains in the new nor­mal. I can­not tell you how many dis­cus­sions with our clients that I have where the VP of sup­ply chain is now respon­si­ble for man­u­fac­tur­ing. Why? Because of the link between intel­li­gence sup­ply chains and smart man­u­fac­tur­ing. It is all about deci­sion sup­port to help busi­ness users. Our tech­nol­o­gy ana­lyst referred to this as ‘aug­ment­ed trans­ac­tions’, which means embed­ding AI and ML into busi­ness soft­ware and work­flows.

“With an inter­con­nect­ed sup­ply chain, com­pa­nies can be pre­dic­tive and pre­scrip­tive around work­flows to dri­ve auto­mat­ed action, such as replen­ish­ing require­ments from dif­fer­ent inven­to­ry loca­tions to meet peak peri­od demand. Auto ML is a new tech­nol­o­gy we’re putting into those work­flows. There are a lot of chal­lenges around skill short­ages in the area of data sci­ence and ana­lyt­ics, so build­ing AI and ML into our sup­ply chain work­flows is more impor­tant than ever.”

For more infor­ma­tion, vis­it intersystems.com/supplychain

Pro­mot­ed by Inter­Sys­tems