Mahesh Babu MG: Pioneering I-AI-Augmented Decision Quality ku-SAP Manufacturing Planning

I-manufacturing global landscape is in a state of perpetual evolution, ebonakalayo umkhosi we-supply chain complex, imibuzo emakethe emangalisayo, futhi ingcindezi eside ye-hyper-efficiency. Amakhasimende eyenziwe yokukhiqiza, elide isisekelo sokukhiqiza, isizinda ngokuvamile ngokuvamile ngokuvumelana nezidingo zokuphendula isikhathi esifanayo kanye nokukwazi ukulawula isigaba esikhulu se-variables. Lo mkhakha lithuthukise isidingo esikhulu sokuqinisekisa izinhlelo zokusebenza. I-AI iyatholakala njenge-power transformative enhle, enikezela ukwandisa kakhulu izinzuzo zokuzimela, ukuphucula ukuxhaswa kwezinto ezincinane, nokushintshiselela izinhlelo zokusebenza zokusebenza kuzo zonke izinzuzo zokukhiqiza. Ngokuvamile, izakhiwo zokukhiqiza ezivamile zihlanganisa izinzuzo ezifana nokuvumelana nezinsizakalo nezinsizakalo kanye nezinsizakalo zokusebenza. Ngaphandle kwalokho, izinhlelo zokusekelwe ku-AI zokusekelwe ukucubungula zingathola ama-algorithms ezihlangene kwedatha e-real-time, ukucubungula ukucubungula kwama-demand ngokunemba kakhulu, futhi ukucubungula imizuzu yokukhiqiza ngokushesha ngezimo ezintsha. Lokhu ukucubungula akuyona kuphela ukucubungula kwezobuchwepheshe, kodwa ukuguqulwa kokuqinisekisa indlela yokukhiqiza nokuphumula. I-Mahesh Babu MG, umongameli obumangalisayo we-SAP Supply Chain Manufacturing, uqala ukuhlanganiswa kwe-AI ku-sophisticated domain ye-SAP Manufacturing Solutions. I-Mahesh Babu MG, umongameli we-SAP Supply Chain Manufacturing. I-Carrier yayo, ebandayo eminyakeni angu-19, ibonise iziphakamiso eningi kanye nenkinga eningi e-SAP Manufacturing and Planning Solutions, ebandakanya i-SAP ECC, i-SAP APO, ne-SAP S/4HANA. I-MH has demonstrated a deep understanding of manufacturing business processes and architectural design across a multitude of industries. Ngokusho kwamanje, uqhuba i-SAP Premium Hub CoE Manufacturing and PLM team, isigaba esihlanganisa umphathi wayo kanye nokuhlola kwezobuchwepheshe emangalisayo. I-SAP S/4HANA Production Planning & Manufacturing Expert, i-MG uye umbhali we-2 editions ye-SAP Production Planning & Manufacturing. , “PP/DS with SAP S/4HANA” (ngokusho, isahluko yokuqala, ISBN 978-1-4932-1872-1, futhi Ngathi I-SAP Press Book Ukubuyekezwa Okuzenzakalelayo izixhumanisi ku-Advanced Planning and Scheduling, ezihlanganisa izindawo ezinzima ezifana ne-master data, i-heuristics, i-PP/DS optimizer, ne-alarm monitoring, ziye zihlanganisa. , ukucindezelwa kwama-status yayo njenge-premise authority ku-field. I-catalogue ye-Biblioteca del Congresso Uhlanganiswa okuzenzakalelayo kwe-SAP PP/DS enhle, enhle kanye ne-forward-thinking research ku-AI augmentation inikeza kuye njengomthombo enhle phakathi kwezinhlelo zebhizinisi ezivamile kanye ne-wave elilandelayo ye-intelligent technologies. Lokhu kuqinisekisa ukuthi iziphumo zayo zibe ezintsha futhi zitholakala ngokushesha ku-real-world manufacturing contexts. Ukubambisana kokubili yakhe esikhulu ku-SAP PP/DS ne-HANA prototyping, u-MG wahlala futhi wahlala isifundo esikhulu se-behavior. Le nophando yenzelwe ukuhlola ngokufanelekileyo izinga le-management decision when operating under AI-augmented planning systems. I-experiment ebonakalise i-AI co-pilot yekhono, eyenziwe nge-capacities ye-SAP Joule, ehlanganisiwe ngqo ne-PP/DS functionalities. Ngezansi zokusebenza kwe-AI co-pilot, abasebenzisi abanolwazi ukufinyelela izixazululo ze-scheduling angu-30% ngokushesha, zibonise isivinini se-20% ephakeme sokubambisana nezinhlelo ezivamile ze-heuristic nge-AI-driven. Imiphumela ye-experiment yaziwa kakhulu: Abasebenzisi abenziwa yi-AI co-pilot angakwazi ukufinyelela izixazululo ze-scheduling ngokushesha kwe-30%, zibonisa isivinini se-acceptance esikhulu se-20% ye-sequences eyenziwe ngempumelelo, futhi, ngokuqondile, ukunciphisa isizinda se-12%. Lezi iziphumo ezinguquguqukayo ze-statistically zihlanganisa ngokushesha ukuthi umphathi we-AI I-concept ye-AI copilots efana ne-SAP Joule, ebonakalayo izibuyekezo zokusebenza ngokuvumelana nezinsizakalo nge-language yemvelo, kuyinto esiyingqayizivele lapha. Nakuba i-Joule ye-standard ye-PP/DS ayikho ngesikhathi sokucwaninga, ukuhlola le paradigm ye-MG ngokusebenzisa imodeli ye-conceptual ikakhulukazi ibonakalayo. Ukuxhumana okuqondile phakathi kokuphendula kwekhwalithi e-AI nokunciphisa i-12% yokushintshwa kwe-late orders ibonise imiphumela yebhizinisi ebonakalayo, okuhlobisa indawo enkulu yokufakelwa ku-manufakili ehlanganisiwe ne-customer satisfaction ne-exploitation costs. Izinzuzo ezitholile ezivela kulezi zophando ziye zithunyelwe ngokuvamile kwamasiko zokusebenza ze-MG kanye nezivakashi zayo ezahlukene. Lokhu kwandisa i-reputation yayo njengomfundisi we-think leader emkhakheni eyenza ngokushesha e-AI-driven manufacturing planning. Umphumela professional we-MG uye lihlanganisa kakhulu ukuncintisana izimboni zokukhiqiza e-North America ukusetshenziswa okwenziwe ngempumelelo izixazululo zokukhiqiza kwe-SAP Supply Chain. Umphumela wokugqibela kuyinto ukuphucula kakhulu izinsizakalo zokukhiqiza kanye nezimo zokukhiqiza zokusebenza zayo. I-industry yokukhiqiza, ikakhulukazi e-North America, isitimela isikhathi esikhulu sokuguqulwa, okwenziwe kakhulu ngokubanzi kwe-digitalization kanye nezidingo esizayo yokwakha amasevisi zokukhiqiza eziningana nezimo. Ukusetshenziswa kwe-AI iyatholakala ngokubanzi njenge-enabled pivotal in this ongoing shift, ne-percentage enkulu ye-companies actively exploring or already implementing AI technologies within their operations. I-MG’s work is directly confronting the practical challenges and unlocks the opportunities inherent in this evolving industrial landscape. I-North America is a significant market for SAP SCM consulting services, nge-multiple large manufacturing enterprises actively seeking specialized expertise to navigate these changes. Umqondo we-"SAP Joule" e-PP/DS ye-MG isifundo linikeza isakhiwo se-proactive, ukuhlola umthamo wesithathu se-AI ngaphambi kokuba izixazululo ze-standard zitholakala emhlabeni jikelele. Lokhu kuyindlela esiyingqayizivele yezinhlangano ezinikezele ukufinyelela ngokushesha kwe-technological advances. Ukuphendula ama-PP/DS ezivamile nge-AI-driven insights I-Production Planning and Detailed Scheduling (PP/DS) izinhlelo ezivamile, nakuba ezinzima, zihlanganisa izinzuzo ezijwayelekile ezihlangene nokuvuthwa kwe-AI augmentation. I-MG ibonise, "Nge-PPDS ezivamile, ama-planners ne-schedulers zihlanganisa kuma-algorithms ye-background execution planning and scheduling kanye ne-optimizers e-PPDS ukuze zihlanganise imibuzo ye-planning ne-scheduling ezinxulumene nokukhiqiza imikhiqizo yokukhiqiza kanye nama-assembly. Nokho, akukho zonke imibuzo ye-planning ne-scheduling angakwazi ukufinyelelwa nge-planning/scheduling algorithms ne-optimizer I-Planer ibonise isikhathi esikhulu ukucubungula ama-alerts eyenziwe nge-transaction interactive, ukucubungula imibuzo efanelekayo yokuthumela ama-orders yokuthengisa ngenxa ye-capacity engabikho ku-assembly line, noma ama-days supply calculations abalandela ngaphansi kwezingcele ezinzima ezinzima. Kulezi zimo, ama-Planer zihlanganisa imizuzu yokukhiqiza noma zihlanganisa imizuzu ezintsha ukunciphisa imibuzo. Ama-methodology yokukhiqiza ezivamile zihlanganisa izinhlelo zihlanganisa njenge-demand variability, ukunciphisa i-real-time operational visibility, kanye ne-inherent inflexibility lapho zihlanganisa izinguqukayo ezintsha, konke oku I-SAP Joule kuyinto i-co-pilot enokufundisa i-semantics yebhizinisi ku-cloud ERP system. Ukuze i-PPDS, akukho izinzuzo ze-Joule ezingenalutho ezisebenzayo kwezixazululo ze-standard. Ngakho, le nkqubo yokufundisa isekelwe ku-impact kanye nezinzuzo ze-co-pilot / chatbot-enabled ukukhiqizwa nokuhlela okuhlobene. " Isicelo esisodwa sokusebenza kuyinto ukuqeqesha imodeli ye-AI usebenzisa idatha yokukhangisa kwe-PP/DS enhle kanye nezinyathelo zokuphendula ezihlobene ezinyathelwe ngama-planner. Lolu hlelo yokusebenza okuzenzakalelayo ukuchofoza izixazululo ze-Alert Monitor ezingama-frequently observed and often time-consuming within the PP/DS system and enhance metrics like times lead and on-time delivery by meticulously considering resource and component availability. Nangona kunjalo, ukusebenzisana kwe-heuristics futhi i-alert monitor yokulawula ama-exceptions, kufuneka ama-planners abahlanganyele ngempumelelo afana ne-"day's supply calculation results in a value that is below a defined threshold" noma "i-delay in delivery of a sales order driven by nonavailability of capacity." I-Vision ye-AI Co-Pilot, enokufanele ukuxhumana nge-context yebhizinisi kanye nokuxhumana nge-language yendalo, isixazululo esisebenzayo esebenzayo esihlanganisa ngezindaba ezivamile. Lezi zokusebenza okuzenzakalelayo zibonisa izinga lokuphumula kakhulu, isizinda esithakazelisayo se-AI emakethe yokukhiqiza. Ukucubungula i-AI nge-planner action, isifundo se-codify ne-scale-scale yobuchwepheshe asebenzayo asebenzayo asebenzayo asebenzayo asebenzayo asebenzayo, ngokuvamile ukuhlangabezana izinhlelo ezinhle kanye nokushesha ukufundisa izigulane ezintsha. Lokhu ukucubungula okuqhubekayo kwezidingo ze-co-pilot, ngisho ngaphandle kwe-PP/DS Joule isisombululo esisodwa, inikeza amathuba esizayo kwezinhlangano ezivela ku-AI. I-pivotal role ye-HANA prototyping ekukhiqizeni i-AI-augmented experiments I-Architecture Technical ye-MG Behavioral Experiment yasungulwa kakhulu ngu-expertise yayo ku-SAP HANA prototyping. Lokhu ukuqeqeshwa kubalulekile ekwenzeni izakhiwo zokuxhumana eziyisisekelo ezidingekayo lokuhlola. I-MG ibonisa, "I-HANA prototyping expertise iyasetyenziswa ekwenzeni izakhiwo zokufaka izici ze-alerts kanye nezinqubo ezisebenzayo kwama-planner e-HANA amamodeli, okuyinto zingasetshenziselwa ngokushesha ukuqeqesha imodeli ye-AI." Lezi zokusebenza, ikakhulukazi izitebhisi kanye nezibuyekezo ze-HANA, zihlanganisa ngokufanelekileyo i-nuances ye-alerts eyenziwe ngaphakathi kwe-PP/DS kanye nemisebenzi yokuguqulwa ezilandelayo eyenziwe ngu-human planners. Le mekanism yokufaka idatha iye iye iye iye iye iye iye yendiza, ukuqinisekisa ukuthi idatha eyenziwe kuyinto ukuguqulwa kwe-dynamic ye-planning activities eqhubekayo. futhi izinzuzo zayo zihlukile ngokushesha ku-reel-time data processing kanye ne-analytics, inikeza isikhungo esiyingqayizivele yokufaka idatha ye-event ye-dynamic esiyingqayizivele yokulungisa imodeli ye-AI enesibophezele. Lokhu kubhalwe ne-trend esikhulu yokusebenzisa i-ML ne-AI ku-SAP data analysis ukuze kuthuthukisa i-reel-time data analysis nokukhuthaza ukucubungula. I-SAP HANA Ukusetshenziswa komkhiqizo we-HANA-based data capture kuncike ngakumbi nge-scenario ezithile MG ibonisa: "Ukuye, ngoba isakhiwo se-Level 3 ku-manufacturing ama-alerts ziye ziye zithunyelwe ukubuyekeza ku-planner ngokushesha, umplanner wahlukanise ngokushesha i-planning heuristics ukuguqulwa kwekhwalithi, njengoba kulesi sikhulu zihlanganisa kakhulu. Ngenxa yalokho, lokhu ukuhambisa, kanye nenkinga le-product/location, ngokubambisana nezinqubo ezisebenzayo, zihlanganiswa ngokusebenzisa izakhiwo ze-HANA, njenge-tables ne-visits ezakushintshwe ngokusekelwe kwezinhlobonhlobo ezivela ku-PPDS. "Lesi isibonelo esifundekayo ibonise ukuthi idatha yokusebenza, kuhlanganise izinhlobo ezithile ze-alarm, ulwazi se-contextual mayelana ne-product ne-location, kanye nezinyathelo ze-resolution ezisebenzayo ze-planner, kungenziwa ngokuvamile futhi zihlanganiswa. Ukukhiqizwa kwe-HANA artifacts kuyinto isicelo esizayo se-data preparation techniques esiyingqayizivele yokufundisa imodeli esebenzayo ye-machine learning futhi isixazululo lwezimfuneko esiyingqayizivele ye-data ephakeme kakhulu yokusebenza kwe-AI enhle. Izinto ezintsha ezintsha, njenge-data extraction ye-PP/DS ye-SAP S/4HANA ezihlangene ne-order details, operation, kanye ne-resource capacity, zihlanganisa ukuthuthukiswa kwebhizinisi ukuze kuholele idatha efana ku-analytical, efana ne-MG prototyped. Ukukhetha ukunikezela i-prototype ku-level 3 assembly nge-planning ama-failures eziningana kakhulu ibonise ukwahlukanisa indawo ye-high-impact lapho i-AI angakwazi ukunikezela inani elikhulu ngokuphathelene ne-permanent pain point, imikhiqizo engcono ye-AI-use-case identification. Ngaphezu kwalokho, ukuthatha akhawunti kanye nemisebenzi kuphela, kodwa futhi umzimba we-product/location inikeza ukwakhiwa kwe-contextual data set enhle. Lezi zebhizinisi ze-multidimensional zihlanganisa ama-AI amamodeli eyenziwe ngempumelelo emangalisayo ngaphandle kwe-alert classifications amancane, okuholela ku-recommendations eyenziwe nge-AI ezincinciphayo futhi ezinxulumene emkhakheni. Ukwakhiwa kwe-comparisons eqinile: Ukulungiswa kwe-participant ne-scenario ye-AI decision support studies Umklamo we-workshop e-MG’s isifundo wahlanganisa kakhulu ukwahlukanisa umxholo we-participant kanye ne-scenario ngokuphathelene ukubambisana okuqinile futhi enhle phakathi kwezimpendulo ezivamile ze-human-driven yama-AI co-pilot. Ngokusho umzimba we-inthanethi we-studium, izinketho ezithile ziye ziye zitholele ukunemba kwezimpendulo. I-MG inikeza lokhu ngokuvumelana, "Kuba isixazululo se-studium etholakalayo futhi etholakalayo ezimbini zezinhlobo ze-alarm (ukuguqulwa kwe-order and over utilization of capacities), izigaba ezidlulile ziye zihlanganisa ukukhiqizwa kwezimpahla ezivamile kanye ne-scheduling supervisor ye-finished goods assembly lines." Lezi zihlanganiswa okuhlobene zihlanganisa ezimbini ze-alarm ezinzima kanye nezinhlelo zokusebenza ze-supervisory ezivela ku-analytics engaphezulu, engaphansi kwe-AI ku-impact ye-finished goods assembly lines. Ukukhetha umphathi wokukhiqiza imikhiqizo ephelele kanye nomphathi wokubeka umphathi wokubeka umugqa wokubeka imikhiqizo ephelele njengama-party roles kuyinto ingcindezi esiyingqayizivele kwenqubo yokusebenza. Lezi zomthetho zihlanganisa ngqo futhi ngokushesha ekulawula izinhlobo zezimali ezijoliswe-ukushintshwa kwe-order fulfillment kanye nokushintshwa kwe-capacity over-use. Lokhu ukucindezeleka kuncike ukuthi izicelo ezakhiwe ngesikhathi lokusebenza zokusebenza zihlanganisa emkhakheni, okungenani zihlanganisa ngokuthandana nezimfuneko ze-real-world kanye nezimpendulo ezivela kumakhasimende. I-scenarios, eyenziwe nge-delay in order fulfillment and overuse of capacities, isixazululo iinkinga ezivamile ezindaweni zokukhiqiza. Lezi zinkinga zihlanganisa kwezimfuneko eziphambili zokusebenza kanye nezimfuneko zokusebenza, okwenza ukunciphisa okungagunyaziwe ngokusebenzisa i-AI isixazululo enhle yezinkampani. Ukusetshenziswa kwezinkulungwane eziholile kuyindlela ephakeme futhi efanelekayo emzimbeni yobuchwepheshe yokuhlanganisa ukusebenza yabantu ngaphansi kwezimo ezihlukahlukene, njenge-like nge-AI kanye ngaphandle kwezixhobo zokuxhumana izixazululo. I-“controlled” nature of these workshops inikeza ukujabulela imiphumela yesifazane ye-AI by minimizing the influence of extraneous variables. Ngaphezu kwalokho, for a robust comparison in studies evaluating human versus AI decision-making, ama-scenarios kufanele Ngokuvamile, izinhlobo ze-alert eyenziwe ngu-MG zihlanganisa izindlela ezinhle noma ezinhle ze-resolution, okuvumela ukufinyelela kwama-benchmark ebonakalayo ukuthi izicelo ezivela kumadokhumenti ezivela kumadokhumenti we-Human kanye ne-AI zihlanganisa. Lokhu ukuhlanganiswa okuhlobene kwezinhlobo ezithile ze-alert nge-rolls ezithile ze-supervisory kuqinisa i-realism ye-experiment kanye, ngokuvamile, ukusetshenziswa kweziphumo zayo kuma-structures ezivamile ze-AI kanye nezinsizakalo zokusebenza. Ukubalwa umphumela: Imikhiqizo ye-Key Metrics for Evaluating AI-Augmented Planning Decisions Ukubuyekeza ngokugqithisileyo imiphumela ye-AI co-pilot, i-MG ibonise ikhasi le-metric esiyinhloko esihlanganisa ngesivinini sokuphendula, izinga le-AI-recommended se-actions se-sequences, kanye neziphumo zokusebenza ezibonakalayo zokusebenza njenge-order timeliness. Ukubuyekeza kanjani okusheshayo izixazululo zingatholakala nge-AI assistive, isifundo lihlanganisa ukuphendula kwe-AI. MG inikeza, "Ukuhlukanisa isivinini kusetshenziselwa ukunambitheka phakathi kwe-input prompt ku-chatbot kuya ku-production ye-response yokuqala. Ngokwesibonelo, ngenxa ye-prompt, 'izinto ezingu-priority-high-delayed order alerts zihlanganisa imikhiqizo echazweni', i-chatbot iyasebenzisa i-AI ukuxhumana ne-prompt kanye nokukhuthaza imisebenzi ze-backend PPDS ukuvelisa ama-alerts kanye nokuphendula nge-number of alerts." U-Metric inikeza ngokuphathelene umthamo we-AI yokusebenza ngokushesha izilinganiso ze-language, ukuxhumana ne-PP/DS yesistimu esifundeni, futhi inikeza ama-planners ulwazi olufanelekayo, umphakeli ebalulekile emkhakheni wokukhiqiza okusheshayo. Lokhu ukuxhumana lihlanganisa izinga lokusebenza ze-AI assistant ezivamile. I-Quality ye-AI sugestions, ikakhulukazi ukuvumelana kwezimpendulo zayo zayo zokusebenza, lithathwa ngokusebenzisa isilinganiso esiyingqayizivele. "Ukuhlolwa kwe-contextual precision isetshenziselwa ukuvumelana kwe-sequence ukuvumelana kwezimpendulo ezivela ku-AI, ngokuvumelana nesimpendulo se-manual ye-actions eyenziwe ngegama elifanayo, "I-MG ithi. Lokhu ukufinyelela engaphezu kokubili okuvumelana okuzenzakalelayo noma ukuchithwa kwegama le-AI. Okokuqala, i-order timeliness, i-KPI ebalulekile yokukhiqiza okwenziwe ngokuvamile nge-KPI yokukhiqiza, iyizimo oluphambili, ebonakalayo ekunciphiseni kwama-12% yama-orders angapheli ekupheleni kwe-angle ye-studium. Imikhiqizo ye-AI Model Validation iyasetshenziselwa ukuqinisekisa ukucaciswa nokunambitheka kwezi ama-metrics, kuhlanganise ukuhlola imibuzo ye-chatbot ngenxa ye-relevance ne-correctness. I-combination ye-metrics – isivinini, izinga le-decision/action sequence, kanye ne-operational outcome – inikeza isakhiwo esihlalweni sokuqinisekisa ukusebenza kwe-AI co-pilot, ukuqinisekisa ukuthi ukuguqulwa kwel I-example prompt ye-decision speed inikeza futhi ukucindezeleka kwe-AI enhle, okwazi ukucindezeleka kwilwimi yemvelo, isixhumanisi yomsebenzisi (isib. "i-products I'm responsible for"), nokukhuthaza ukuxhumana kwamakhemikhali e-backend emangalisayo. Lokhu kubonisa izinzuzo eziphambili ze-planning assistants emkhakheni. I-Dynamics Unforeseen: Ukuxhumana kwe-Planer nge-AI suggestion vs. i-heuristics yemvelo Ukusebenzisana phakathi kwama-planners abantu kanye nezinhlelo zokuxhumana zokusebenza kwe-AI kunokukwazi ukucubungula ama-dynamics ezingathandeki, futhi isifundo se-MG inikeza iziphakamiso ezinhle kuleli khasi. Enye yezimpendulo eziyinhloko zihlanganisa ukucubungula okuphumelela okuphumelela kwe-training data ku-AI amayunithi. I-MG ibonisa, "Nge-Joule-like modeled chatbot, ikhwalithi kanye nenkqubo yebhizinisi idatha kubaluleke kakhulu izixazululo ze-AI. Kwiimeko ze-capacity overload, idatha yebhizinisi lihlanganisa ama-scenaries lapho ama-orders ezidlulile i-overload zithunyelwe kwiveki elilandelayo." Ukulungiselela ukuqeqeshwa okuqala kubangelwa kwesimo se-AI esifanele, kodwa engaphansi kwe-optimum. Ukubaluleka kwe-Quality ye-Data ye-Training kuyinto isisekelo esiyingqayizivele ekuthuthukiseni i-AI; imiphumela emibi angakwazi ukuhlaziywa okungafanelekayo nokuhlukaniswa okungenani. Ukulungiselela kwama-patterns ezivamile ezivela ku-scenario enhle. I-MG ibonise, "Kuma-capacity iyatholakala ngaphakathi kwiveki elifanayo, i-AI ibonise ukuguqulwa kwe-order ku-semana elandelayo; lokhu lithathwe ngokuvumela ukuguqulwa kwe-capacity njenge-action ngaphambi kokuvumela ukuguqulwa kwe-action ku-training data." Lezi zokusebenza kwe-AI, lapho ibonise isixazululo esizayo ngenxa ye-biases ekusebenzeni kwayo, ibonise ukuthi amamodeli we-AI angakwazi ukufundisa imibala ezingenalutho uma idatha ye-training ayikho ephelele kakhulu noma ibonakalayo zonke imibuzo yokwenza imibuzo. Ukusebenza okufanayo, uma akuyona, kungabangela ukucindezeleka kwama-planner futhi kuholele izixazululo zokusebenza ezingenalutho. I-strategy yokunciphisa – ukucubungula idatha ye-training ukuze kuhlanganise ukucubungula kwe-capacity ngokuvumelana – ibonise inqubo yokuthuthukisa okuqhubekayo ebonakalayo ekuthuthukiseni imodeli ye-machine learning. Lolu hlelo lihlanganisa ukuhlangabezana "ukudlulisela" noma izinsizakalo ezivamile ezivamile ezisetshenziswe ngempumelelo, njenge-checking yokufinyelela kwezinto ezingenalutho ngaphambi kokufunda ukuchithwa okwengeziwe. I-AI amamodeli akufundisa kusuka kuma-patterns, futhi uma idatha ye-training akufanele ngokugcwele yonke isixazululo se-decision logic, i-AI uzothola izixazululo zayo. I-AI wahlanganyela isipho se-suboptimal (ngokusho ukuhlolwa kwezimpendulo zebhizinisi) kuyinto ebalulekile ukuguqulwa kwalo, okukhuthaza i-value ye-transparency ne-explainability emakhemikhali ye-AI, okuyinto ezinzima ukwakhiwa kwamakhasimende etholakalayo nokuvumela ukuphathwa kwamakhemikhali efanelekayo. Lolu hlobo le- iterative, lapho isivumelwano se-AI ikhiqiza ukuxhumana kwamakhasimende amakhulu kanye nokuvamile ukuxhumana okuphazamiseka ngenxa ye-AI amaphuzu, ivumela ukuqinisa kwe-AI, okuyinto kubalulekile ekuthuthukiseni izindlela zokuxhumana kwamakhemikhali e-AI ezisebenzayo futhi asebenzayo. Yini Izici ezibonakalayo: Imiphumela yokusebenza nezimali sokunciphisa amahora nge-AI Ukunciphisa ama-12% emaphaketheni amancane e-MG ye-conceptual study ibonise izinzuzo eziningi kakhulu nezinhlangano zezimboni, zihlanganisa izinga lokusebenza kanye nezinzuzo zebhizinisi. Izindlela ezivamile zokuphendula ama-order delays, ikakhulukazi zihlanganisa izixazululo ze-store-floor, zihlanganisa kakhulu futhi zihlanganisa. Njengoba MG inikeza, "Ngama-scenarios ezivamile yokuhlanganisa, uma ukuhlaziywa komthengisi yokuthengisa kubangelwa kwezimpendulo zokukhiqiza imikhiqizo ngexesha elidlulileyo, izivakashi zangaphakathi ziye ziye ziye ziye ziye zithunywe ukuze zihlanganise izicelo zokukhiqiza zangaphakathi ku izikhwama zokusebenza zokusebenza zokusebenza. Emva kwalokho, ukucubungula ukucubungula izicelo zokuthengisa ziye ziye ziye ziye ziye zithunywe ukuze zihlanganise izinsuku ezintsha zokuthengisa izikhwama." Ngaphezu kwalokho, ukulawula imizamo yokuthengisa ephezulu-priority ngokuvamile kufuneka ukwelashwa okuzenzakalelayo kumadivayisi ukuguqulwa kwezinye imizamo yokukhiqiza, inqubo eside futhi okungenani ingxubevange. I-AI-augmented chatbot approach, njengezijoliswe ku-studium, inikeza ukuguqulwa kwe-paradigma ku-resolution enhle futhi e-automated. I-MG ibonise lokhu ukucindezeleka: "Nge-chatbot-based approach, umsunguli angakwazi ngokuphathelene u-chatbot ukuba ezinye izicelo zokuthengisa ze-high-priority zihlukile futhi zitholele ukuguqulwa inkinga, okuyinto ngokuvumela amandla ze-AI ukukhuthaza umsebenzi wokuguqulwa okuqhubekayo ngokuvumelana ngokuvumelana ngokuvumelana ngokuvumelana." Lezi zokusebenza ngokuvumelana ngokuvumelana nesisekelo esikhulu se-inefficiency ne-cost. Ukunciphisa ama-12% emaphaketheni amancane ibhizinisi kubuyekezwa kakhulu. Ngokwemvelo, imithombo emaphaketheni emaphaketheni futhi emaphaketheni emaphaketheni ibhizinisi ngokushesha emaphaketheni kanye namahora emaphaketheni, ukunciphisa ukunakekelwa kwamakhasimende, okuholela ukunciphisa, futhi kungabangela ngisho ukuphazamiseka. Ngakho-ke, ukunciphisa amaphaketheni amaphaketheni ukwandisa ukusetshenziswa kwamakhemikhali futhi imithombo yokukhiqiza okuqinile. Izimpendulo zekhwalithi zihlanganisa ngokulinganayo, ne-potential for direct cost savings (umthombo we-on-time we-poor can account for 10% of the costs), i-augmented revenue from enhanced customer loyalty, and healthier profit margins. Research kusuka kumazwe afana ne-Aberdeen Group has shown a correlation between higher on-time delivery rates and improved project profitability, with best-in-class firms significantly outperforming others in this metric. Ngaphezu kwalokho, i-automation yokuguqulwa kwe-charting ye-critical, i-high-priority orders akukwazi ukunciphisa kuphela ukuphepha lokusebenza okuqondile we-planner, kodwa, okungenani kuncike ukuthi izici zebhizinisi ze-strategic, njenge-satisfaction ye-customer orders, zihlangene ngokuhambisana nokushesha. Ngokwenza lokhu, ukunciphisa ingozi yemali ehlanganisiwe nezi-akhawunti ezinzima, kubonisa ukuthi i-AI inokukwazi ukuvikela ama-inkokhelo ezinzima kanye nokukhuthaza imibuzo eziyinhloko zebhizinisi ngokuvumelana ngokuphathelene ne-prioritized kanye ne-effective handling of urgent fulfillment challenges. Ukuguqulwa kusuka ku-reactive, i-batch-oriented re-scheduling kuya ku-proactive, i-AI-driven resolution inikeza ukwandisa kakhulu ku-planning agility. Lokhu kukuvumela abaphakeli ukuguqulwa ngokushesha futhi ngokuvumelana nokunciphisa imiphumela yayo ye-cascading ye-supply chain. Ukusuka ku-experiment kuya ku-expertise: Ukuhlobisa i-executive workshops ne-think leadership nge-AI insights Imiphumela enhle ye-behavioral experiment ye-MG, ikakhulukazi izinzuzo ezibonakalayo ekusebenzeni kwe-planner kanye nokukhula kwekhwalithi lokuphendula, ziye ziye ziye zihlanganisa izakhiwo zokusebenza zokusebenza kanye nezitolo zokusebenza zayo, kuhlanganise nezithombe zayo ze-SAP Press. Umfundisi wokuqala wesixhumanisi wesixhumanisi nabasebenzi be-industry kuyinto umthamo wokuguqulwa kwe-AI yokuvumela i-"manage by exception" paradigm e-management planning. Njengoba i-MG ibonise, "I-planner ye-typical noma i-scheduler ikakhulukazi ama-thousands of materials ku-area yayo ye-responsibility, futhi isikhathi yayo yokuthumela ukucubungula ama-planning/capacity alerts e-PPDS ezingenalutho." Ukusabela okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe okuqukethwe. Ukukhiqizwa kwe-AI-driven tools inikeza indlela esifanele yokunciphisa le ibhizinisi. I-MG inikeza ukucindezeleka okuphakeme kwama-AI: "Uma unemibuzo yayo nokufakwa ku-alerts emangalisayo kanye nokulawulwa kwebhizinisi. Lokhu ukucindezeleka kwe-AI-driven yokukhiqiza inikeza i-business kanye ne-planning team yayo ekuphakeme ekuphakeme ku-"Ama ukulawula uhlelo, kuphela ukulawula imiphumela yayo." Lezi philosophy, lapho izinhlelo zokusebenza kwe-AI ngokufanelekileyo ukuhlangabezana ama-alerts ezivamile nezimpendulo ezincinane, i-human planners inikeza ubuchwepheshe kanye ne-bandwidth ye-cognitive yokuvakashela izinhlelo ezinzima kakhulu, ukulawula ama-exceptions eziyinhloko zebhizinisi, kanye nokufinyelela izinhlelo zokusungula zokusebenza. Umgangatho we-”management by exception” iyisisombululo esidumile ku-operations managers Izici ze-MG, njenge-30% yokushintshwa okusheshayo nokunciphisa ama-orders angu-12%, inikeza izitifiketi ezithile, ezokufundisa idatha, okuhlobisa ukuhlolwa okuhlobene kwezimpumelelo ze-AI ku-demonstrations ezinhle ze-value. Lezi zokusebenzisana okuholela ikakhulukazi ku-executive workshops, okuholela abacwaningi ekuphenduleni kwezimpumelelo zayo ze-AI. t Ukusebenza kwe-MG, ngakho-ke, akudingeka kuphela ukulungiselela kwedolobha, kodwa kusebenza njenge-instrument esebenzayo yokuchofoza nokufundisa, ukunciphisa isixazululo phakathi kwamandla we-AI ne-executive understanding kanye ne-buy-in. Lokhu ukuhlaziywa kwegama we-planner - ukusuka ku-system operator kuya ku-strategic problem solver - kuyinto ingxenye ebalulekile yokuxhumana kwe-AI, okuyinto kufuneka izinzuzo ezintsha kanye nokuguqulwa kwe-mindset ye-organizational, izinto ezinzima ku-MG ye-educational outreach. Ukulungiswa kwe-strategic: Ukuqondisa amazwe ku-AI-augmented SAP PP/DS environments Ngokusekelwe iziphumo zophando wakhe, i-MG inikeza iziphathimiso zokusebenza nezinhlangano zokusho ukuhlanganiswa kwama-AI-augmented planning systems ku-SAP PP/DS izindawo zayo zangaphakathi. Isifundo sokuqala, kodwa ephakeme, ibonise izinzuzo ezizodwa zebhizinisi. Njengoba i-MG ibonise, "Nge le ucwaningo oluphakeme nokuhlola, izinzuzo zebhizinisi lithatholakala ku-12% ukwandisa imizuzu yokusebenza kwamakhasimende ngokushesha." Ukuphumelela okuhlobisa kusebenza njenge-startpoint enhle futhi isitifiketi se-AI. Ngokusekelwe lokhu, isixazululo esiyingqayizivele kuyinto ukuqala ukusetshenziswa kwe-AI ngokuchofoza imiphumela emikhulu yama-use cases lapho i-return on investment iyatholakala futhi iyatholakala. Lokhu kubhalwe ne-inthanethi ezingaphezu kwe-AI yokufakelwa okukhuthaza ukuqala nge izici ezizodwa zebhizinisi kanye nezicelo ezijwayelekile ezibonakalayo ezinikezwayo. MG inikeza izindlela ezithile zokuthuthukiswa okungenani: "Uma izicelo ezilandelayo, njenge-analyzing i-planning and optimizer job logs ukucacisa ama-planning and scheduling errors kanye nokuthuthukisa i-AI ukucubungula ukucubungula ama-errors ezivamile, iyathuthukisa inani elikhulu kwebhizinisi." Ukuhlola i-planning kanye ne-optimizer job logs kuyinto isinyathelo esilandelayo esilandelayo esiyingqayizivele ngoba zihlanganisa ama-logs ezimbonini, ama-structured data sources ezokufinyelelekayo emzimbeni we-SAP. Ukubuyekezwa kwedatha okuqukethwe kwebhizinisi zokusebenza kwe-AI kungenzeka ukunciphisa ukufinyelela kwebhizinisi elandelayo se-AI, okunikezela indlela efanelekayo yokwandisa izinzuzo ze-AI. Ukusebenza okuzenzakalelayo ukucubungula amafakelo ezivamile ezaziwa ku-logs zihlanganisa i-”management by exception” philosophy. Lokhu akuyona kuphela ama-planners, kodwa futhi ukunciphisa umsebenzi we-IT support kanye nama-system administrators, okuvumela ukuba zihlanganise ngezinkinga ezinguqile ze-system noma ukuthuthukiswa, ngokuvamile ukwandisa imiphumela yokusebenza kwe-AI ngaphandle kwe-planning department. Ukuphumelela kwe-12% yokuphumelela kwe-time fulfillment kusebenza njenge-proof-of-concept ebalulekile. Ukukhuthaza ukuhlangabezana nezinsizakalo ezilandelayo zihlanganisa i-strategy yokubonisa inani elifanelekayo ngokushesha bese ukusetshenziswa kwalo ukuchitha nokuphumelela kokuphumelela kwe-AI emangalisayo ngaphakathi kwenkampani – isisekelo esisodwa yokulawula ukuguqulwa okuphumelela kanye nokuphumelela kwe-strategic. Izinkampani kufanele ukunikeza izinga lokufaka kwedatha kanye nokulawula, njengama-pre-requisites eziyinhloko zokuphumelela kokuphumelela kwe-AI ku-SAP emkhakheni. Nangona ukunakekelwa kwe-MG lihlanganisa ekuphumelela kwezinsizakalo z Umhlahlandlela wokusebenza kwe-AI ku-manufacturing planning, njengezithombe nokusebenza kwe-MG, ibonise indlela yokukhuthaza amandla zokusebenza okuphumelela kakhulu. I-behavioral experiment, ebonakalisa ukuthi i-AI co-pilot yekhompyutha ye-SAP PP/DS ingathuthukisa i-decision speed nge-30% futhi ukunciphisa ama-orders e-latent nge-12%, inikeza izici ezinzima lokuphumula kwe-AI. Ngaphezu kwalokho, lezi ziphumela ukuguqulwa kwegama le-planner kuya ku-"management by exception" model. Ngokusebenzisa ukucubungula ukucubungula ama-alerts ezivamile nezimo ezivamile, i-AI inikeza abacwaningi abacwaningi ukunikezela izinzuzo zabo zokusebenza nezinhlangano ezinguqile ezingenalutho. Lesi-synergy phakathi kwe-intelligence yabantu ne-artificial intelligence ayikho inkqubo esithakazelisayo enhle, kodwa kuyinto enhle, ukunikezela zezimboni zokukhiqiza, ikakhulukazi ezivela e-North America, ukusebenzisa izixazululo ze-SAP, isitimela lokuphumula okuphumula, ukucubungula okuphumula, kanye nokuphumula okuphumula. Ukuhlolwa kwama-pioneer kanye nemibuzo yestrategic ezinikezelwe abacwaningi afana ne-MG zihlanganisa lokhu ukucubungula, ukuqinisekisa ukuthi ukusetshenziswa kwe-AI iyatholakala kanye nokuvumelana nezidingo ezintsha ze-modern manufacturing. Ukuhlolwa okuqhubekayo kwe-Intelligent Manufacturing Planning ngeke ikakhulukazi kuhlobisa lokhu ukusebenzisana okuqhubekayo. Lapha, ukulawula umzimba ukulawula amandla we-AI ukuvikela izinga ezintsha zokusebenza nezinsizakalo.