CUT­TING EDGE AU­TOMA­TION IM­MI­NENT

Get ready for Rock­well Au­toma­tion’s en­try into sub-Sa­ha­ran Africa

The Star Early Edition - - BUSINESS REPORT - Andile Ma­suku is a broad­caster and en­tre­pre­neur based in Jo­han­nes­burg. He is the ex­ec­u­tive pro­ducer at AfricanTechRoundup.com. Fol­low him on Twit­ter @Ma­sukuAndile and The African Tech Round-up @african­roundup Andile Ma­suku

THE NEW YORK Stock Ex­change-listed firm, Rock­well Au­toma­tion, is the world’s largest provider of in­dus­trial au­toma­tion prod­ucts. Rock­well em­ploys more than 22 000 peo­ple and has cus­tomers in more than 80 coun­tries world­wide.

How­ever, the Wis­con­sin-based com­pany, which ped­dles both the Allen-Bradley and Rock­well Soft­ware brands among oth­ers, main­tains a fairly low pro­file out­side its area of spe­cial­ity and has a rel­a­tively mod­est foot­print in sub-Sa­ha­ran Africa.

To ac­cel­er­ate growth across the con­ti­nent, Rock­well ac­quired Hiprom in 2011. Hiprom, a Jo­han­nes­burg-based cor­po­ra­tion, is a lead­ing process con­trol and au­toma­tion sys­tems in­te­gra­tor spe­cial­is­ing in min­ing and min­eral pro­cess­ing.

When Rock­well ac­quired Hiprom, a com­pany spokesper­son re­vealed that the ac­qui­si­tion was a strate­gic play to strengthen their global project man­age­ment and de­liv­ery ca­pa­bil­i­ties in the min­ing, met­als and min­er­als in­dus­tries.

Ac­cord­ing to John Lewis, Rock­well’s cur­rent di­rec­tor of busi­ness part­ner­ing, Hiprom – which is still run out of South Africa – is now the group’s global min­ing com­pe­tency cen­tre of ex­cel­lence.

In a re­cent pod­cast con­ver­sa­tion I had with Lewis, he shared how Rock­well is adapt­ing to chang­ing times by hir­ing soft­ware de­vel­op­ers and tech-savvy busi­ness spe­cial­ists who can speak to the myr­iad of op­ti­mi­sa­tion chal­lenges faced by their clients all over the world.

When asked what per­cent­age of Rock­well’s out­put, in terms of the so­lu­tions they de­liver to clients, is hard­ware ver­sus con­sult­ing ser­vices and soft­ware, Lewis stated that the ra­tio is roughly 70 per­cent hard­ware, 20 per­cent en­gi­neer­ing ser­vices and 10 per­cent soft­ware.

He has­tened to add that the mix is chang­ing rapidly, mov­ing away from hard­ware and grow­ing to­wards so­lu­tions and soft­ware be­cause of to the broad global trend to­wards dig­i­tal trans­for­ma­tion.

When Lewis first joined Rock­well in 1979, the com­pany was al­most 100 per­cent a hard­ware busi­ness and their ser­vice propo­si­tion to fac­tory own­ers was “Buy our stuff, and we’ll come out and re­place any­thing that breaks”.

Over time, that con­cept grew to in­clude: “En­list us to help you en­gi­neer so­lu­tions.”

Lewis ad­mits their clients had a hard time ad­just­ing to be­ing charged for con­sult­ing ser­vices, but ap­par­ently soon enough re­alised the ben­e­fits of hav­ing a com­pe­tent tech­ni­cal part­ner on call to keep ma­chine down­time to a min­i­mum.

As com­put­ing played a more sig­nif­i­cant role in how lead­ing in­dus­tri­al­ists hacked op­er­a­tional ef­fi­ciency is­sues, soft­ware de­vel­op­ment and de­ploy­ment be­came more and more im­por­tant.

To­day, Rock­well is ex­pe­di­tiously re­search­ing and test­ing var­i­ous soft­ware ap­pli­ca­tions and ser­vice de­liv­ery mod­els which of­ten in­volve the de­ploy­ment of ar­ti­fi­cial in­tel­li­gence (AI), ma­chine learn­ing and the In­ter­net of Things (IoT).

When asked how he re­sponds to pro­labour crit­ics who as­sert that com­pa­nies like Rock­well un­der­mine liveli­hoods by help­ing in­dus­tri­al­ists har­ness au­toma­tion to com­pletely elim­i­nate the need for hu­man par­tic­i­pa­tion in fac­tory pro­cesses, Lewis stated that he is yet to en­counter a “lights out, no hu­mans in­volved” in­dus­trial op­er­a­tion.

He reck­oned that it is largely un­safe and oner­ous tasks, as well as repet­i­tive jobs which are dif­fi­cult for hu­mans to do con­sis­tently, that are be­ing au­to­mated.

Unique skills

Lewis in­sisted that while all such work is be­ing taken over by ma­chines, many other jobs are be­ing cre­ated re­quir­ing dif­fer­ent and more unique skills.

He did, how­ever, ad­mit that such jobs are not nec­es­sar­ily cre­ated at a rate of one for one, ref­er­enc­ing the grow­ing need for in­di­vid­u­als pos­sess­ing higher tech com­pe­ten­cies to in­stall, pro­gramme and main­tain cut­ting-edge in­dus­trial equip­ment, as well as write and in­te­grate soft­ware.

Babusi Ny­oni is a Zim­bab­wean se­nior user ex­pe­ri­ence (UX) de­signer at Thom­son Reuters and is based in Cape Town.

Ny­oni hap­pens to be low-key, but one of the con­ti­nent’s lead­ing AI and ma­chine learn­ing prac­ti­tion­ers.

In Oc­to­ber 2016, he gave a TEDx talk on how pre­dic­tive mod­el­ling and his­toric data could be used to an­tic­i­pate Africa’s next refugee cri­sis.

Shortly after­wards, the UN Refugee Agency (UNHCR) in Geneva reached out to rope him in as a con­sul­tant.

Since then, he has helped the UNHCR build a pro­to­type that uses con­flict and food se­cu­rity data to pre­dict the mag­ni­tude of dis­place­ment in one of the world’s war-torn na­tions.

A pro­duc­tion ver­sion of the tool is al­ready in the works, and once it is ready, the UNCHR plans to make it fully ac­ces­si­ble for use by gov­ern­ments, civic or­gan­i­sa­tions, cor­po­ra­tions, and in­di­vid­u­als look­ing to pre-empt im­pend­ing hu­man­i­tar­ian crises.

I have come to value Ny­oni’s views on how ad­vances in ro­bot­ics and au­to­mated soft­ware are likely to change ev­ery­day life, not least be­cause we share a fairly ide­al­is­tic world view.

Dur­ing a re­cent in­ter­view, Ny­oni told me that de­spite spend­ing a great deal of time work­ing on re­tail AI ap­pli­ca­tions, he is most ex­cited about the fu­ture of AI in biotech.

He cited how the be­gin­ning of 2017 saw the ap­proval of the first US Food and Drug Ad­min­is­tra­tion-ap­proved ap­pli­ca­tion of ma­chine learn­ing and deep learn­ing for di­ag­nos­ing heart con­di­tions. (Yes, there is a dif­fer­ence be­tween ma­chine learn­ing and deep learn­ing, but I will not be div­ing into that.)

Not only is Ny­oni ex­cited by in­no­va­tions such as the cur­rent use of AI-led com­puter vi­sion to help vis­ually im­paired peo­ple per­ceive the world around them, but he is es­pe­cially en­livened by the prospect of fast-learn­ing soft­ware be­ing de­ployed in ail­ing bod­ies.

As the likes of Rock­well con­tinue to pro­mote the trend to­wards mech­a­nised au­toma­tion across the world’s lead­ing in­dus­tries, the po­ten­tial use cases for IoT will un­doubt­edly mul­ti­ply.

As that hap­pens, we should ex­pect a spike in the de­mand for AI and ma­chine learn­ing ap­pli­ca­tions that will be used to make sense of the vast amounts of data that con­nected de­vices col­lect.

Ul­ti­mately, our need to per­form ac­cu­rate big data anal­y­sis needs to keep up with such ad­vances if mankind is to ben­e­fit from IoT de­ploy­ment. Ny­oni reck­ons that if we fail at this, the con­se­quences could be cat­a­clysmic – pic­ture hun­dreds of thou­sands of pace­mak­ers mal­func­tion­ing, fac­to­ries melt­ing down and hun­dred-car pile­ups.

What Ny­oni and I most de­cid­edly do not have in com­mon is his Elon Musk-es­que view that hu­man­ity is speed­ily edg­ing to­wards a “sin­gu­lar­ity with the ma­chine”. He would point to the way so­cial me­dia foot­prints are be­com­ing an ex­ten­sion of peo­ple’s ex­is­tence as op­posed to an al­ter­nate plane, as might have been the case ini­tially.

Out­sourced

Ny­oni be­lieves that be­cause many of us have out­sourced de­ci­sion-mak­ing power to AI’s such as Google’s to in­form how we nav­i­gate our daily lives, ie in­ter­act with fel­low hu­mans, re­late to our phys­i­cal en­vi­ron­ment, and plan for the fu­ture, we might be open­ing our­selves up to cat­a­strophic events should the AIs we’ve al­lowed to “run our lives” be com­pro­mised.

On one hand, I to­tally dis­card the very no­tion of sin­gu­lar­ity.

On the other, I take John Lewis’ as­ser­tion that in­dus­trial au­toma­tion won’t com­pro­mise liveli­hoods with a mas­sive pinch of salt, par­tic­u­larly within the con­text of the de­vel­op­ing world.

The liveli­hoods de­bate aside, I do think that there are far more com­plex is­sues we would all do well to stay awake to as busi­ness in­ter­ests look to em­brace ro­bot­ics and ex­ploit soft­ware au­toma­tion.

I be­lieve that as cut­ting-edge au­toma­tion tech­nolo­gies con­verge, the per­fect storm is im­mi­nent. And be­fore it hits, we ought to de­cide what kind of hu­man be­ings we want to be.

Shall we pas­sively al­low new age in­dus­tri­al­ists free rein to pur­sue any profit-driven au­toma­tion projects they wish, or should we lobby for the com­plete democrati­sa­tion of his­toric data cur­rently held by pro­pri­etary en­ti­ties and in­sist that any in­no­va­tions launched in the au­toma­tion space be judged solely on the ba­sis of tan­gi­ble pub­lic ben­e­fits such as im­prove­ments in health­care de­liv­ery and food pro­duc­tion?

PHOTO: SUP­PLIED

Babusi Ny­oni is a se­nior user ex­pe­ri­ence (UX) de­signer at Thom­son Reuters.

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.