The role human workers play in automation

Often talk of automa­tion is met with some side­ways glances and whis­pers of jobs being made redun­dant. How­ev­er, if the last few years have shown us any­thing, it is that some­one needs to be there to fix the self-check­out in the local super­mar­ket.

Essen­tial­ly, automa­tion helps to relieve the pres­sure on work­ers dur­ing a labour short­age or at peak times, or remove the need to man­u­al­ly per­form repet­i­tive jobs. This results in stream­lined oper­a­tions and enhanced effi­cien­cy, which helps gen­er­ate more prof­it across the entire busi­ness.  

The impor­tance of automa­tion has grown in the face of Brex­it and coro­n­avirus, both of which promise con­tin­ued dis­rup­tion, labour short­ages and fluc­tu­at­ing con­sumer demand. This, cou­pled with chang­ing cus­tomer expec­ta­tions and a shift online, means chief infor­ma­tion offi­cers (CIOs) need to reassess their automa­tion offer­ing.

Accord­ing to Kofax vice pres­i­dent of brand and cor­po­rate mar­ket­ing John Lipsey: “As organ­i­sa­tions con­tin­ue to embrace dig­i­tal work­flow trans­for­ma­tion, automa­tion becomes increas­ing­ly impor­tant. In fact, full-scale end-to-end automa­tion has become a mat­ter of sur­vival in the wake of the pan­dem­ic. More and more cus­tomers demand fric­tion­less expe­ri­ences and busi­ness­es that don’t pro­vide them may not be able to sur­vive.”

Human touch

Despite men­tion of full-scale automa­tion, this does not mean “no humans allowed”. In fact, automa­tion tech­nolo­gies work best when used along­side humans and should be a key fac­tor for CIOs dur­ing the imple­men­ta­tion process. 

“When applied smart­ly, tech­nol­o­gy can help pro­vide relief for human work­ers by facil­i­tat­ing repet­i­tive jobs. This can gen­er­ate lee­way and the poten­tial for more through­put,” says ProGlove chief exec­u­tive Andreas Koenig. “The main ben­e­fit of this is that work­ers can then focus on the tasks they do bet­ter than machines.”

How­ev­er, when it comes to imple­ment­ing automa­tion, it’s cru­cial to know where it is going to be best placed. “CIOs need to keep in mind that automa­tion has its lim­its,” says Koenig. “Some process­es just can­not be auto­mat­ed for a num­ber of rea­sons, includ­ing too many prod­uct vari­ants dur­ing many man­u­fac­tur­ing process­es.”

For exam­ple, in the ecom­merce mar­ket, which has seen tremen­dous growth even before the pan­dem­ic. “Increased ecom­merce trans­ac­tions impact the ful­fil­ment process as they come with dras­ti­cal­ly dif­fer­ent deliv­ery times and for­mats that will need to be han­dled. One cus­tomer may order a tooth­brush, anoth­er a TV set. These two orders can­not be deliv­ered through the same lines,” says Koenig.

“Automa­tion comes with a sub­stan­tial invest­ment, lim­it­ed appli­ca­tions and is time con­sum­ing in the deploy­ment process too. So it is absolute­ly crit­i­cal to pre­cise­ly iden­ti­fy the oppor­tu­ni­ties and ben­e­fits automa­tion may or may not deliv­er. Pri­ori­ti­sa­tion and knowl­edge of the restric­tions of the tech­nol­o­gy to be deployed are also crit­i­cal aspects for CIOs to con­sid­er.

“Aside from ade­quate tal­ent, all this calls for thor­ough knowl­edge of the organ­i­sa­tion and its process­es. So it may make sense to build that tal­ent inter­nal­ly. And even though it may sound trite, it is manda­to­ry to pro­vide guid­ance, prop­er change man­age­ment and clear com­mu­ni­ca­tion when intro­duc­ing automa­tion.”

Getting the foundations right

For Koenig, the key tech­nolo­gies to look out for are “opti­cal char­ac­ter recog­ni­tion, intel­li­gent char­ac­ter recog­ni­tion, nat­ur­al lan­guage pro­cess­ing, robot­ic process automa­tion, machine learn­ing or deep learn­ing”. 

But don’t pan­ic if you haven’t even thought twice about some of these tech­nolo­gies because the needs of each busi­ness will vary across indus­tries. CIOs need to be mind­ful that there is not a tick-box list of automa­tion tech­nolo­gies they need to have and, as Koenig points out, there is no sil­ver bul­let or a one-size-fits-all approach.

A major ben­e­fit of automa­tion is the lev­el of scal­a­bil­i­ty it offers, there­fore start­ing with the basics can help solid­i­fy a good foun­da­tion for fur­ther invest­ment, as well as grad­u­al­ly intro­duc­ing the work­force to new ways of oper­at­ing. 

“At the end of the day, organ­i­sa­tion­al suc­cess depends on the sup­port of the work­force. Sup­port requires trust and trust is built on trans­paren­cy or the democ­ra­ti­sa­tion of work­flows to sup­port work­ers,” says Koenig.

“It is para­mount to iden­ti­fy the poten­tial of the tech­nol­o­gy you are con­sid­er­ing. Tech­nol­o­gy must help to design bet­ter work­flows and pro­vide relief for tedious, repet­i­tive or mun­dane jobs. In short, it will need to serve human needs. So any deci­sion to auto­mate needs to be dri­ven by and rely on valu­able data.”

First comes data, then comes automation

If busi­ness­es are going to start some­where, they should start with their data, says Koenig.  “Busi­ness­es, even though they need to be com­pli­ant with pri­va­cy reg­u­la­tions, def­i­nite­ly need to work more with the data they cap­ture. Data refine­ment is a key con­cern here. It is crit­i­cal to invest in tech­nol­o­gy that goes beyond sim­ply col­lect­ing infor­ma­tion by con­tex­tu­al­is­ing it so you can deduct action­able insights,” he says.

At the core of automa­tion tech­nolo­gies lies a vast amount of data, col­lect­ed by inter­net of things devices. How well you can process all the data gath­ered will gov­ern the qual­i­ty of busi­ness insights that are gen­er­at­ed.

Accord­ing to Kofax and Qlik’s white paper Using data for insights is still crit­i­cal for busi­ness oper­a­tions: “As the world evolves and more oper­a­tions con­tin­ue to be dig­i­talised, busi­ness insights from data con­tin­ue to be one of the fastest and most effec­tive ways to derive val­ue, from short-term actions to long-term invest­ments.” Once the data has been cap­tured, “com­plex­i­ty and errors can be removed from this resource inten­sive, man­u­al process through automa­tion”.

Fur­ther ben­e­fits can be derived by lever­ag­ing automa­tion to com­bine “busi­ness data with data from oth­er dis­parate sources and chan­nels – cus­tomer data, cus­tomer feed­back, oper­a­tional data and pro­cess­ing data – and pro­vide sim­pler, smarter ways to visu­alise busi­ness intel­li­gence”. But the real trick, accord­ing to Kofax and Qlik, “is tak­ing all the data acquired through intel­li­gent automa­tion and turn­ing it into action­able infor­ma­tion”.

The white paper con­tin­ues: “Most busi­ness­es have become adept at col­lect­ing lots of data. How­ev­er, find­ing the most effi­cient and effec­tive means of under­stand­ing and democ­ra­tis­ing the data, and using it to make informed busi­ness deci­sions, is more dif­fi­cult for organ­i­sa­tions.”

Cre­at­ing more trans­paren­cy around the avail­abil­i­ty of data and mak­ing it more acces­si­ble can help solve this. “Organ­i­sa­tions and users need self-ser­vice, inter­ac­tive web-based dash­boards that pro­vide process and busi­ness insight, with­out the need to involve IT to build new reports or adjust data­base queries. It’s key to cre­ate cus­tom dash­boards or mod­i­fy exist­ing ones, or to fil­ter and aggre­gate data for dis­play­ing con­tent in a vari­ety of views,” it says.

Powerful trio

COVID-19 has accel­er­at­ed the use of automa­tion with­in pro­cure­ment and sup­ply chains due to its poten­tial for improved cost effi­cien­cy and pro­duc­tiv­i­ty, accord­ing to GEP’s 2021 Sup­ply Chain and Pro­cure­ment Out­look Report. Automa­tion will help stream­line process­es and allow smarter col­lab­o­ra­tion in core pro­cure­ment and sup­ply chain activ­i­ties.

“Automa­tion boosts out­put and affords employ­ees more time to focus on strate­gic work, allow­ing com­pa­nies to do more with less,” the report says. “Plat­forms with built-in fore­cast­ing and sup­ply net­work capa­bil­i­ties will lever­age arti­fi­cial intel­li­gence (AI), machine learn­ing and robot­ic process automa­tion to reduce their reliance on indi­vid­ual spread­sheets and boost shared, real-time data and demand mod­els.”

In anoth­er white paper, Arti­fi­cial intel­li­gence and its impact on pro­cure­ment and sup­ply chain, GEP says: “Apply­ing machine learn­ing to the dis­cov­ery of pat­terns in sup­ply chain data – the pain­points and most cru­cial fac­tors for smooth, suc­cess­ful man­age­ment of a net­work – can be rev­o­lu­tion­ary.

 “As the data is inter­ro­gat­ed under con­straint-based mod­el­ling, the machine can extract the most influ­en­tial cri­te­ria that impact inven­to­ry, demand, pro­duc­tion plan­ning, risk and logis­tics man­age­ment, and sup­ply chain opti­mi­sa­tion.”

Many of the insights gen­er­at­ed may have not have been pre­vi­ous­ly under­stood or even known due to the sheer vol­ume of data avail­able. Access to such insights enables smarter plan­ning deci­sions and intro­duces new lev­els of net­work agili­ty and respon­sive­ness, help­ing busi­ness­es stay ahead of the curve.

Is AI a cure-all?

How­ev­er, the GEP 2021 out­look report also empha­sis­es that automa­tion is not a panacea: “It must be cou­pled with human insight into cat­e­go­ry demand and the impact of a lin­ger­ing reces­sion, and con­tin­ue to man­age fluc­tu­a­tions in safe­ty for employ­ees and part­ner sup­pli­ers.

“In atyp­i­cal sit­u­a­tions, though, algo­rithms alone can fall short. Ear­ly on in the pan­dem­ic when cus­tomers were sud­den­ly request­ing toi­let paper instead of LEGO, Amazon’s algo­rithms were per­plexed. The sit­u­a­tion demon­strates that bulk order­ing and extreme demand surges need human inter­ven­tion to revise ful­fil­ment and inven­to­ry replen­ish­ment.

“Busi­ness­es that com­bine human adapt­abil­i­ty with machine-learn­ing mod­els will bet­ter with­stand the shock of unfore­seen dis­rup­tions. Algo­rithms still need a dai­ly real-world con­text.”

Like­wise, when it comes to imple­ment­ing AI, it helps to start small and spe­cif­ic. GEP sug­gests CIOs look to tar­get spe­cif­ic prob­lems. This allows teams to  build con­fi­dence and by tak­ing the time to under­stand where the biggest val­ue can be extract­ed, and then deliv­er­ing a short­er project with well-defined objec­tives, this acts as a good test­ing ground before expand­ing the AI reach.

---