More smartness, more value
Achieve more sustainable value with built-in and premium value enhancers. Based on AI, data science, deep learning and cognitive features, they solve specialist business-critical challenges to make smart plans even smarter.
Collaboration and simulation
Allocation and promising
Solvers and ML (AI)
Machine learning is great, but is it going to crush statistical forecasting?
May 3, 2021
Machine learning is the talk of the town but is it going to crush statistical forecasting completely? The recent success of machine learning in the M5 competition has inspired some...
machine learning is great but it going to crush statistical forecasting value enhancers the talk of town completely recent success in m5 competition has inspired some commentators predict death i think they are cutting corners while rightfully claiming its place demand there good reasons believe that even many years from now will be used conjunction with as a complementary rather than substitute technique here&rsquo s why   forecasting: extrapolating well-defined set data let me first briefly summarize principles technology all about using techniques such regression sampling and hypothesis testing on historical time series sales allows trends seasonal patterns correlations external factors identified extrapolated based assumption past behavior best predictor future frequently holt-winters model for example essentially extrapolates average values cyclical figure below illustrates what does showing blue red here’s supply chain planning neural network you don&rsquo t need master&rsquo degree statistics understand covid-19 pandemic severely disturbed 2020 into 2021 which undermines their predictive forecasters apply several additional smooth out or otherwise adjust before them compute forecast ml solve problem not at just like uses including where was undermined by virus old gigo principle garbage applies much learning: identifying broader range use these different way presents opportunities put simply links inputs outputs there&rsquo no define equation upfront feed huge amounts carefully selected input output perhaps shipment large number products intervals then identify linked this so-called phase during trained &lsquo wires&rsquo immediately highlights one biggest appeals ml: can relatively easily almost modeling effort add same product family master calendar promotional orders point-of-sale weather economic identifies somewhat ingenuously if you&rsquo re feeding right make smarter over downsides mean end have suggested after ml&rsquo so fast only particular dataset we&rsquo sure winning models would also work well other datasets example: lot yet long history limited portfolio simpler probably still your bet heavier computational requirements take days train specialized hardware inherently less transparent more black box rarely big might resistance organization adopting because explained most importantly always guaranteed lead better accuracy adage valid: often cases simple moving provide fancy complement each potential guarantee naive machines sometimes see none vulnerable disrupted it&rsquo very likely come neither do without human intelligence that&rsquo another story want know how help! get touch
What dating apps and supply chain solvers have in common
February 8, 2021
At conferences, my colleague Kris talks a lot about dating. Before you come away with the wrong impression, he’s not a professional dating guru and he doesn’t give out tips. Kris is one of the heads of OMP’s optimization department.
His point is...
dating apps and supply chain solvers what have in common value enhancers at conferences my colleague kris talks a lot about before you come away with the wrong impression he&rsquo s not professional guru he doesn&rsquo t give out tips is one of heads omp&rsquo optimization department his point smart solutions making them understandable for users pursuit now commonly known as xai uses analogies takes deployment planning solver an example this assigns stock volumes to demand incorporating range constraints kpis by using specialists call &lsquo weighted greedy fractional matching algorithm ' may sound either very complicated boring or both but underlying mechanisms that make work are pretty simple much more tantalizing subject online explain it audience its simplest form   imagine female male analogy goes like you&rsquo re building app your aren&rsquo so looking partner they just want fulfill their annual date requirements none lgbtq let&rsquo say women men because #feminism means instead 52 tons we&rsquo jane who wants keep existential loneliness bay having dates year 4 paul only each single has certain preferences when comes potential these could be shared interests appreciation same kind humor similar tastes night movies correspond such transportation costs difference between due availability among many others pairs how alike read blog understand from perspective easily explained first thing do calculate well possible pair fits together take sum over all preference characteristics instance important her can laugh rather than also likes political thrillers higher weighting calculation gives match representing good fit would percentage on okcupid then sort matches way down list start best oliver bouldering comedian daniel sloss time 15 considers almost calling hobby oliver&rsquo get assigned no we ignore other pairings sorry girls there&rsquo enough go around still need remaining 37 she game moving second-best it&rsquo kate five william 45 william&rsquo excluded rest understanding logic prerequisite trust third-best case happy every saturday &mdash continue until men&rsquo given women&rsquo been satisfied works exact finding values quantities sorting working metaphor easy see really allows step user results if were eager deploy data science machine learning curious our browse through faqs submit own question learn
Supply chain planning language for robots (and humans)
November 9, 2020
The world of supply chain planning sometimes appears to us as a vast, ever-changing field of knowledge, with new concepts being introduced constantly. Recent additions to the vocabulary are terms like ‘telescopic digital twin’, ‘agile planning’,...
supply chain planning language for robots and humans value enhancers the world of sometimes appears to us as a vast ever-changing field knowledge with new concepts being introduced constantly recent additions vocabulary are terms like &lsquo telescopic digital twin&rsquo agile planning&rsquo resilient xai&rsquo but what do they mean mathematician i&rsquo m big fan clear definitions simple examples illustrate complex that&rsquo s ll try here first thing that should probably be clarified is none these refer any specific technology there isn&rsquo t one single algorithm behind xai example or generates rather seen underlying philosophies when it comes choice design technologies   twin so let&rsquo say you building robot sidekick help manage your after you&rsquo ve given cute face little wheels roll around on good friend&rsquo name ann might explain looks going need gartner defines this following: &ldquo representation physical can used create plans make decisions&rdquo * means inside mighty brain re construct mental image including all plants machines transport lanes well detailing how interact fit together zooming in out an read blog find explainable artificial intelligence makes if equipped she put information into context asking her opinion whether open plant pawnee indiana will then able look at picture&rdquo decision sense than getting caught up details similarly making decisions more detailed level sequence tasks machine day allows zoom relevant practice omp done by modeling using multiple interlinking twins corresponding different layers decision-making scopes adapting itself also always needs see changes puts it: must synchronized real just model &rdquo ties concept agility refers easily picked tell adding don&rsquo want say: hold wasn&rsquo explained layout me could please whole again instead quickly adapt some systems take even further allowing system dynamically detect inaccuracies suggest appropriate our  continuous data improvement  track called genie adapts accurately uses learning techniques calculate change rates based production avoiding disruptions now knows suggestions equip smart solver algorithms ai designed such way unexpected have massive effect resulting plan our basically sure too skittish doesn&rsquo completely overhaul once something happens achieved combination forecasting analysis methods predict place buffers dampen their effects understand has computed optimal - tells wonder got conclusion ask reply cause trust where acronym idea advanced come set tools allow them themselves putting abstract calculations back shows interconnected ideas work aware scope therefore each kept mind designing effective deserve robin batman kitt knight rider leslie future explore road map innovation insight 16 october 2020 tim payne
The role of XAI in supply chain
October 26, 2020
How can people learn to trust the systems that are supposed to make their lives easier?
That was one of the big issues discussed at the recent 'Smart People in the Smart Supply Chain' virtual conference hosted by Vlerick Business School in partnership with OMP, bluecrux and...
the role of xai in supply chain value enhancers how can people learn to trust systems that are supposed make their lives easier was one big issues discussed at recent 'smart smart chain' virtual conference hosted by vlerick business school partnership with omp bluecrux and lineas experts from all over world dialed for an afternoon talks workshops presented leading voices both academia industry overarching topic 4 0 &mdash or technologies like ai other sophisticated algorithms be harnessed shape future planning   humans machines human-machine is a very important issue its most benign distrust leads avoidance sector this might result reluctance implement new destructive generate aversion even fear these really helping us they about steal our jobs read blog find out explainable artificial intelligence solvers talk entitled &ldquo we trust&rdquo professor karlien vanderheyden cristina danila janssen pharmaceutica conjectured lack caused combination three aspects: human factors such as users having unclear expectations system machine transparency what do contextual arising insufficient poor management communication eliminate it&rsquo s clear have figure let explain themselves ways understand appreciate ann vereecke ghent university kris dockx referred concept answer challenge according consists set tools techniques help not only but also interpret recommendations were made model &rdquo does work then went on demonstrate brings life consultants use three-step approach getting end complex solver behind automatic optimization plans: first step explaining algorithm works theory because procedures assign stock demand sound quite dry boring explanation more engaging achieved comparing it something everyone finds way interesting case online dating so instead assigning based abstract kpis explains through device lonely men single women hair color preference two practice user sees every move makes why video animation being able correct when wrong told which moves dislike learns take into account next time therefore becomes intuitively understandable controllable relationship between happy any things essential: understanding best relationships worked will look explore road map
Interview: Harmonizing human and artificial intelligence
April 9, 2020
Q&A with Philip Vervloesem, SVP OMP USA for Supply Chain Management Review.
Is it realistic to fully robotize supply chain planning?
"The level of automation in supply chain planning has tremendously increased and I strongly believe that it will...
interview harmonizing human and artificial intelligence interview: value enhancers q&a with philip vervloesem svp omp usa for supply chain management review is it realistic to fully robotize planning "the level of automation in has tremendously increased i strongly believe that will continue do so both on the demand side ai technology used achieve this however a touch still essential you need automate regular routine tasks provide planners strong analytics they can optimize closed-loop plan also helps ensure all levels&mdash from strategic operational&mdash are synchronized carefully balancing priorities concerns stakeholders including business finance commercial at we call unison planningtm approach "   how your customers benefit managing their global streams "ai vertical horizontal integration network full visibility takes people out silo-based thinking example by showing locations inventory allows anticipate sense limit disruptions maximize outsmart competition see what should cut or push profit proves useful dealing major events such as covid-19 crisis there&rsquo s an important ux component not only smart deploys way it&rsquo easy present outcomes that&rsquo because want show executives which scenario give best market share growth read about robotized artifical impact roles team members "planners adopt different mindset beyond role operations planner think more globally better understand interdependencies actions decisions through simulations have real end-to-end view company-wide collaboration becomes solid cross-functional processes players be successful ai-driven help teams reach consensus does optimally "unison services no coincidence advisory  user engagement services  among our fastest growing service areas upfront turns projects right direction engagement  includes hands-on coaching make smarter shows them discover innovation