![Lokad](/img/default-banner.jpg)
- 289
- 501 209
Lokad
Registrace 6. 09. 2011
Quantitative supply chains are the future: where every possible future gets a probability, where every possible decision gets an economic score, and where control is given back to management thanks to extensive automation. Lokad's goal is to make this future a reality. We help companies do more with their supply chains, their inventories and their production capacities than classical methods would ever allow. Our technology combines big data and machine learning. Our teams deliver the supply chain expertise.
Lokad: l’Intelligence Artificielle au Service de la Supply Chain
Grâce à son expertise de pointe dans le domaine de la supply chain et à l’apport de technologies de dernière génération faisant appel à l’IA, la société Lokad aide les entreprises à améliorer l’efficacité de leur chaîne d’approvisionnement. Présentation.
Une expertise Homme-machine pour optimiser la supply chain.
À travers son activité, la société Lokad aide de nombreuses entreprises françaises et internationales à booster et optimiser leurs supply chains. Pour y parvenir, les équipes de l’entreprise s’appuient sur la combinaison de l’analyse des données permises par le differentiable programming - un descendant du deep learning -, le cloud computing, ainsi qu’un haut niveau d’expertise dans le domaine de la supply chain.
Pour enrichir les analyses de données, Lokad a également recours aux dernières avancées dans le secteur de l’intelligence artificielle, notamment avec les LLM (grand modèle de langage). Chaque client bénéficie également d’un accompagnement complet de la part des experts de l’entreprise : les Supply Chain Scientist.
« L’objectif est de proposer une réelle expertise Homme-machine pour permettre à nos clients d’optimiser leur supply chain, que ce soit pour gérer le dimensionnement de leur stock, pour le placement de mouvements de stock au sein de leur réseau, mais aussi pour affiner leur stratégie de prix et de promotion, ou pour leur production », explique à ce sujet Estelle Vermorel, co-fondatrice de l’entreprise.
Aider les entreprises à prendre de meilleures décisions
Lokad qui travaille aujourd’hui aux côtés de nombreux acteurs issus de secteurs d’activité divers et variés - aéronautique, e-commerce, grande distribution, etc. -, réalise au quotidien un travail d’innovation constant, afin de continuer à proposer des solutions technologiques de dernière génération au service de la supply chain.
« Dans le secteur de la supply chain, le terme de résilience est très important. Toutes les décisions s’avèrent être des compromis soumis à de très nombreux risques. L’objectif pour Lokad est donc de réussir à intégrer correctement tous les risques qui sont inhérents à la supply chain, pour aider nos clients à prendre les meilleures décisions ! » conclut Joannes Vermorel, co-fondateur de l’entreprise.
******
Check out our website: www.lokad.com/
Follow us on LinkedIn: www.linkedin.com/company/lokad/
Read our blog: blog.lokad.com/
******
Une expertise Homme-machine pour optimiser la supply chain.
À travers son activité, la société Lokad aide de nombreuses entreprises françaises et internationales à booster et optimiser leurs supply chains. Pour y parvenir, les équipes de l’entreprise s’appuient sur la combinaison de l’analyse des données permises par le differentiable programming - un descendant du deep learning -, le cloud computing, ainsi qu’un haut niveau d’expertise dans le domaine de la supply chain.
Pour enrichir les analyses de données, Lokad a également recours aux dernières avancées dans le secteur de l’intelligence artificielle, notamment avec les LLM (grand modèle de langage). Chaque client bénéficie également d’un accompagnement complet de la part des experts de l’entreprise : les Supply Chain Scientist.
« L’objectif est de proposer une réelle expertise Homme-machine pour permettre à nos clients d’optimiser leur supply chain, que ce soit pour gérer le dimensionnement de leur stock, pour le placement de mouvements de stock au sein de leur réseau, mais aussi pour affiner leur stratégie de prix et de promotion, ou pour leur production », explique à ce sujet Estelle Vermorel, co-fondatrice de l’entreprise.
Aider les entreprises à prendre de meilleures décisions
Lokad qui travaille aujourd’hui aux côtés de nombreux acteurs issus de secteurs d’activité divers et variés - aéronautique, e-commerce, grande distribution, etc. -, réalise au quotidien un travail d’innovation constant, afin de continuer à proposer des solutions technologiques de dernière génération au service de la supply chain.
« Dans le secteur de la supply chain, le terme de résilience est très important. Toutes les décisions s’avèrent être des compromis soumis à de très nombreux risques. L’objectif pour Lokad est donc de réussir à intégrer correctement tous les risques qui sont inhérents à la supply chain, pour aider nos clients à prendre les meilleures décisions ! » conclut Joannes Vermorel, co-fondateur de l’entreprise.
******
Check out our website: www.lokad.com/
Follow us on LinkedIn: www.linkedin.com/company/lokad/
Read our blog: blog.lokad.com/
******
zhlédnutí: 89
Video
ISF2024 Panel: Demand Planning and the Role of Judgment in the World of AI/ML
zhlédnutí 390Před měsícem
Full transcript available: www.lokad.com/tv/2024/7/17/demand-planning-human-judgement-and-ai/ Panel background The panel was first proposed by Robert Fildes (Professor Emeritus, Lancaster University) in response to Conor's [article critiquing FVA](/forecast-value-added/). This article was republished in the Q2 2024 edition of Foresight (produced by The International Institute of Forecasters, th...
Optimizing Omnichannel Purchase & Allocation at Worten (with Bruno Saraiva)
zhlédnutí 190Před 2 měsíci
Full transcript available: www.lokad.com/tv/2024/7/8/optimizing-omnichannel-purchase-and-allocation-at-worten/ Summary: Filmed onsite in summer 2024 at Worten’s flagship store in Lisbon, Bruno Saraiva (Head of Stock and Space Management at Worten Portugal) discusses Worten’s ongoing supply chain revolution. Bruno explains to Conor (Head of Communication at Lokad) how the companies are optimizin...
Supply Chain Debate - Is S&OP a net good for companies? (with Milos Vrzic)
zhlédnutí 630Před 2 měsíci
Full transcript available: www.lokad.com/tv/2024/6/19/is-sop-a-net-good-for-companies/ Summary In a debate hosted by Conor Doherty of Lokad, Milos Vrzic, former Head of Supply Chain (EMEAC) at Galderma, and Joannes Vermorel, CEO of Lokad, discussed the value of S&OP for companies. Vermorel critiqued S&OP as simplistic and outdated, while Vrzic emphasized its role in tactical planning. The debat...
Probabilistic Forecasts & Sequential Decision-Making (with Warren Powell) - Ep 163
zhlédnutí 906Před 3 měsíci
Full transcript available: www.lokad.com/tv/2024/5/29/probabilistic-forecasts-sequential-decision-making/ Summary In a recent LokadTV interview, Conor Doherty, Joannes Vermorel, and guest Warren Powell discussed probabilistic forecasts and decision making in supply chains. Warren Powell, a retired Princeton professor and Chief Innovation Officer at Optimal Dynamics, shared his career journey an...
Decision-making under Uncertainty in Supply Chain (with Dr. Meinolf Sellmann) - Ep 162
zhlédnutí 593Před 3 měsíci
Full transcript available: www.lokad.com/tv/2024/5/22/decision-making-under-uncertainty-in-supply-chain/ About the guest Dr. Meinolf Sellmann is founder and CTO at InsideOpt, a US-based startup that produces general-purpose software for automating decision-making under uncertainty. He is the former Director for Network Optimization at Shopify, Lab Director for the Machine Learning and Knowledge...
Forecast Congruence and Supply Chain Decision-Making (with Nikolaos Kourentzes) - Ep 161
zhlédnutí 409Před 3 měsíci
Full transcript available: www.lokad.com/tv/2024/5/16/forecast-congruence-and-supply-chain-decision-making/ About the guest Nikolaos Kourentzes is a professor in predictive analytics and AI at the University of Skövde AI Lab in Sweden. His research interests are in time series forecasting, with recent works in modelling uncertainty, temporal hierarchies, and hierarchical forecasting models. His...
Quality-Cost Dilemma in Supply Chain - Ep 160
zhlédnutí 261Před 4 měsíci
Full transcript available: www.lokad.com/tv/2024/4/24/quality-cost-dilemma-in-supply-chain/ In a dialogue with Lokad's Head of Communication, Conor Doherty, Lokad CEO Joannes Vermorel discusses the quality-cost ratio in supply chain management. Vermorel emphasizes that quality refers to decision-making, not product attributes, and that client-perceived quality may not align with optimal supply ...
Effects of AI on Supply Chain Jobs - Supply Chain in 3 minutes
zhlédnutí 868Před 5 měsíci
Conor Doherty of Lokad scrutinizes a Harvard study on AI's impact on white-collar jobs, revealing nuanced effects. The research, involving 758 consultants, assesses AI's role in enhancing productivity, particularly in supply chain management. It finds that AI boosts performance in certain tasks, especially with training, but may falter in complex scenarios. Doherty critiques the study's narrow ...
Rethinking S&OP (The Future of Supply Chain) - Ep 158
zhlédnutí 995Před 5 měsíci
Full transcript available: www.lokad.com/tv/2024/3/13/rethinking-sales-and-operations-planning/ Lokad CEO Joannes Vermorel critiques the Sales and Operation Planning (S&OP) process as outdated and inefficient for modern businesses. He argues that S&OP, designed for simpler times, struggles to keep pace with today’s complex, fast-moving business environment. Vermorel criticizes the process’ slow...
Stocks are not in your control - Ep 157
zhlédnutí 477Před 6 měsíci
Full transcript available: www.lokad.com/tv/2024/3/6/stocks-are-not-in-your-control/ In this LokadTV eposide, Conor Doherty interviewed Joannes Vermorel on inventory planning misconceptions. Vermorel dismantled the fallacy that stock levels are a direct lever for client satisfaction and profitability. He argued that companies should focus on serving clients profitably, not on the illusion of st...
Skills for Modern Practitioners - Supply Chain in 3 minutes
zhlédnutí 533Před 6 měsíci
Conor Doherty, Head of Communication at Lokad, emphasizes the evolution of supply chain management, highlighting three critical skills for future practitioners: technical writing, digital literacy, and financial awareness. Technical writing is crucial for documenting complex processes, ensuring knowledge is preserved and shared. Digital literacy, encompassing coding and data analytics, is vital...
Stochastic Optimization of Supply Chain Decisions - Ep 156
zhlédnutí 1,1KPřed 6 měsíci
Full transcript available: www.lokad.com/tv/2024/2/21/stochastic-optimization-of-supply-chain-decisions/ In a discussion between Lokad's CEO, Joannes Vermorel, and Head of Communication, Conor Doherty, the importance of stochastic optimization and probabilistic forecasting in supply chain management is emphasized. Vermorel explains the concept of stochasticity, where the loss function is uncert...
A day to day exhibition of Lokad usage in a MRO company (with Rodrigo Pineda)
zhlédnutí 210Před 6 měsíci
Note: The audio quality changes at [7:07] due to unforeseen technical difficulties. We apologize for any inconvenience and thank you for your understanding. Full transcript available: www.lokad.com/tv/2024/2/14/a-day-to-day-exhibition-of-lokad-usage-in-a-mro-company/ In a conversation revealing the intricacies of aviation MRO supply chain management, Baptiste Miceli of Lokad and Rodrigo Pineda ...
RFP Madness in Enterprise Software - Ep 154
zhlédnutí 226Před 7 měsíci
RFP Madness in Enterprise Software - Ep 154
Resilience, Risk, and Effective Leadership in Supply Chain (with Knut Alicke) - Ep 153
zhlédnutí 320Před 7 měsíci
Resilience, Risk, and Effective Leadership in Supply Chain (with Knut Alicke) - Ep 153
Discussion with Joannes Vermorel at the Ecole des Mines de Paris (French)
zhlédnutí 457Před 8 měsíci
Discussion with Joannes Vermorel at the Ecole des Mines de Paris (French)
Large Language Models in Supply Chain (with Rinat Abdullin) - Ep 152
zhlédnutí 970Před 8 měsíci
Large Language Models in Supply Chain (with Rinat Abdullin) - Ep 152
The Role of Scarcity in Supply Chain - Ep 151
zhlédnutí 178Před 9 měsíci
The Role of Scarcity in Supply Chain - Ep 151
MRO Holdings' Quantitative Supply Chain Revolution (with Ricardo Alvarez Henao)
zhlédnutí 290Před 9 měsíci
MRO Holdings' Quantitative Supply Chain Revolution (with Ricardo Alvarez Henao)
Quantitative Supply Chain: What our clients think
zhlédnutí 135Před 9 měsíci
Quantitative Supply Chain: What our clients think
The Evolution of Supply Chain Education (with Paul Jan) - Ep 150
zhlédnutí 305Před 9 měsíci
The Evolution of Supply Chain Education (with Paul Jan) - Ep 150
Risks in Supply Chain Management - Ep 149
zhlédnutí 265Před 10 měsíci
Risks in Supply Chain Management - Ep 149
Supply Chain Board Games (with Mathias Le Scaon) - Ep 148
zhlédnutí 254Před 10 měsíci
Supply Chain Board Games (with Mathias Le Scaon) - Ep 148
Forecast Value Added - Supply Chain in 3 minutes
zhlédnutí 666Před 10 měsíci
Forecast Value Added - Supply Chain in 3 minutes
Guided Tutorial of Lokad's Demonstration Account
zhlédnutí 774Před 10 měsíci
Guided Tutorial of Lokad's Demonstration Account
Customizing Trek Bikes (Mastering Configurability with Dan Scharneck) - Ep 147
zhlédnutí 181Před 11 měsíci
Customizing Trek Bikes (Mastering Configurability with Dan Scharneck) - Ep 147
Pilotage des stocks chez Celio avec Julie Schaf
zhlédnutí 201Před 11 měsíci
Pilotage des stocks chez Celio avec Julie Schaf
MRO Complexity Explained (Paris Air Show 2023)
zhlédnutí 349Před rokem
MRO Complexity Explained (Paris Air Show 2023)
Anderson Sharon Davis Deborah Young Eric
Lee Kenneth Martinez Eric Thomas Sharon
Hall Charles Lewis Richard Robinson Donald
Johnson Jeffrey Miller Elizabeth Taylor Eric
Williams Maria Davis Joseph Gonzalez Donald
Anderson Paul Davis Michael Walker Scott
Gonzalez Jennifer White Shirley Jackson Barbara
Lopez Joseph Williams Michael Garcia Robert
Clark Jason White Melissa Walker Mary
Miller Joseph Jones Cynthia Anderson Steven
Martinez Robert Robinson Richard Miller David
Davis Barbara Lee Larry Miller Scott
White Jason Walker Jennifer Robinson Cynthia
Brown Matthew Wilson Eric Moore Jose
Rodriguez Paul Young Brenda Moore Sandra
Anderson Frank Thompson Nancy Jackson Maria
Johnson Linda Harris Christopher Lopez Dorothy
Dear Friends, I have a question: Can DDMRP run both sale orders and sale forecast, right?.
According to the DDMRP book (2019, Ptak and Smith), this method governs the inventory buffers mostly through a direct application of the ADU (average daily usage). The book remains fuzzy on how this ADU is computed exactly, but simple averages are somewhat implied through most of the examples given in the book. Hope it helps. Cheers, Joannes
Congratulations Bruno🎉!
Any book you would recommend to understand the workings of the supply chain from beginner to pro level?
I am currently working on such a book - precisely because I am quite dissatisfied with the present state of affairs of the supply chain literature. It's going to be somewhat lengthy book.. Stay tuned! Cheers, Joannes
Meanwhile your North American market is failing miserably. We’re throwing away $20,000+ worth of materials away every month. Empower MX is the worst system I have laid my eyes on. Pretty soon you won’t have employees to even work on the customers planes.
While modeling a S&OP system is there any easy way to tackle innumerable constraints in say production processes, customer priorities, technical difficulties etc. or do we have to map each one piece by piece into the algorithm?
In order to mechanize the decision-making processes, there is no alternative, but to "map each one piece by piece into the algorithm" as you very correctly suggest. This is one of the core responsibilities of the Supply Chain Scientists at Lokad. Cheers, Joannes
It would be amazing if you could start adding this to Spotify! 😀
Thanks for providing this format. I've enjoyed viewing it very much. One crucial thing in my opinion is when it comes to automated-decision making there are algorithms which produce "bad" decisions and algorithm which produce "good" decision. Unfortunately the assessment good/bad may come months or even years after the decision. While human communication is slow you may have strong counterparts in the arguments. So major weaknesses can be (theoretically) found early. When it comes to modelling I haven't seen a error free implementation. So my question is: How can you achieve reviews of automatically made decisions with respect to the correctness of the underlying assumptions in the modelling.
The issue of not being able to access the adequacy of a decision until long *after* the decision was made is equally present for people. Having a human making the decision does change anything with regards to this challenge. Humans have no special powers in this area. If at the time the decision is generated (by an algorithm), a human can object to the long-term viability of the decision, and if this objection is well-reasoned; then the algorithm must be modified to immediately take this into account. Automation doesn't preclude keeping human intelligence around to adjust or improve the automation itself. Furthermore , if the decision is automated, it is possible *after the fact* to modify the logic so that the same mistake isn't made again. This property is in sharp contrast with employees who may or may not comply, who may or may not learn from the mistake, or who may arrive fresh and ignorant due to turnover. Joannes
I am working in Order Management System under Supply chain management and I am doing online masters in Data science. I have already 7 years of experience in order management system. Is doing masters in DS right decision in my career?
Becoming proficient in SQL and Python is key. Those are foundational skills that will only become more important in the next decade. Fancy data science is the cherry on the cake, but it won't hurt either. A traditional 'masters' feels very expensive time-wise to gain those skills, but assuming you can complete your online master without quitting your current job, and assuming the price is reasonable, yes, it will certainly put you in the bracket of the more desirable sort of employees. Hope it helps! Cheers, Joannes
Wow, inspiring, I cannot wait to read some of his books now (I’ll start with the tiny URL) and see how I can apply it to my company’s problems.
Amazing talk @Joannes and @Meinolf! I keep learning more from you guys than any conventional books. Is Lokad hiring?
Thank you for the kind word, it's very appreciated. Yes, Lokad is growing very nicely and hiring accordingly. The open positions are listed at www.lokad.com/about-us/#positions Also, you can check out the jobs that we post on LinkedIn. Best regards, Joannes
Amazing talk. Thank you!
Glad you enjoyed it! Joannes
In one of my job as Supply Chain practicioner we build a process where we would look at differences between forecasts (one week vs previous week), trying to understand: * why did the forecast change for this product or product category or country or... * decide if we need to change our plans based on this change much more efficient than recalculating a full distribution plan every week
Absolutely. Yet, not all numerical recipes are equal in their capacity to generate results that make sense - even for the data scientists who understand the algorithms. This is why we have a whole process referred to as 'white boxing' at Lokad to address this. Best regards, Joannes
Which one would you or the guest recommend R or python? I prefer R as it is specifically built for statistics and its functional language is more approachable to someone with a math foundation than OOP of python. However most job listings require python for data science and such.
Python is more versatile option than R, and over the last decade, it has grown into a "classic" general purpose programming language, much like Java or C#. I would recommend Python, but more importantly, I would recommend learning about software and software engineering in general. Mastering the programming syntax is one of the easiest parts of software. Best regards, Joannes
so how should a Business Administration student prepare for this future? Is everything we are studying a waste of time? On Coursera, I am personally studying OR, data science and R programming, although only for statistical decision making and modeling, not for programming. But your videos make it seem like even mathematical programming/linear programming, Statistical modeling and forecasting (prescriptive and predictive analytics) are wastes of time because software like yours and others will do this for us. Furthermore I have been sorely disappointed with Chatgpt R code generating abilities and data analysis. The supply chain scientist role on your website is a domain knowledge expert with data science skills. How can students like myself become SCM experts if your software is gonna make 90% of SCM jobs disappear. College and online courses, even masters degrees can only take one so far. but we will lack any work experience to become those so called domain experts. Some advice for this student would be greatly appreciated. Perhaps some succinct videos for Business student/industrial engineer/SCM or Non software engineering professionals. P.S. I read one of your bitcoin cash blogs a couple years ago and just recently discovered your website. I dont believe POW is 51% attack proof and we have empirical evidence on smaller chains. POS seems to be more resistant like Avalanche.
🤦 "promosm"
what is frontier though...i didn't get???
Hi! It simply means a line/border separating two (or more) things. It is the same term used to describe borders ("frontiers") that separate countries - e.g., the Pyrenees form a natural frontier between France and Spain. In this context, Harvard Business School suggests there is a digital frontier between the things gen-AI (ChatGPT-4) can do well and the things it cannot do well. We disagree. See this essay for a greater explanation of our position: www.lokad.com/blog/2024/4/8/a-nuanced-perspective-on-jagged-technological-frontier/
What we need to keep in mind is that LLMs are not the ideal solution for analytical tasks - though of course they excel in descriptive outputs. We are still waiting to see what Q Star will be in terms of quantitative skills and capabilities - that may be truly transformative...and not in a good way for long term professional employment opportunities for large numbers of people... GPT-4 is still a bit limited in many respects, but future iterations of it, Claude et al will be displaying true leaps in capability. Just hope the travelling salesman/routing problem can finally be easily solved.
It is certainly unfair to use an LLM for complex quantitative tasks and then say "hey, look at how badly it did!" In case you are interested, we expanded our critique of the paper here: www.lokad.com/blog/2024/4/8/a-nuanced-perspective-on-jagged-technological-frontier/
Thank you!
Great talk! Was wondering what’s your take on an aggregate level SOP without going into each and every product but focusing on product families?
Thanks for the kind word! We are going to publish soon an upcoming debate about S&OP. Tons of further materials in this long interview. Stay tuned! Best regards, Joannes
Fabulous! A must for Inventory Management teaching. Thank you!
Thank you, @Lokad. For the amazing content.
Thank you. Best, Joannes
To learn how to code your own supply chain solutions, visit the links below: Envision workshop 1: docs.lokad.com/gallery/workshop-supplier-analysis/ Envision Workshop 2: docs.lokad.com/gallery/workshop-sales-analysis/
Excellent and very informative video! For talks that delve into math, can we also potentially get a reading list that Joannes or Lokad recommends to up and coming supply chain scientists?
Working on it! This is exactly the sort of question that I want the Lokad chatbot to cover. See lokad.com/chat It's not there yet, but I have started to compile a list of book reviews to be (later) fed to this chatbot. Cheers, Joannes
Awesome podcast. Thanks for the content!
The question I have is, where are these 90% of back office workers gonna get new jobs? Looking at these kinds of jobs on the Bureau of Labor Statistics website, most of them only require a high school diploma.
The job market will sort it out, it always does - unless misguided state interventions prevent it do so. New and better jobs will emerge, even if it's unclear what those jobs will be exactly. 150 years ago, farming was +80% of the labor workforce, in the US, in France and pretty much everywhere. Now, it's about 1.5% of the workforce. 90% of the back-office tasks of the 1970s have already disappeared. Remember the time when most junior white collars would spend a few months in the mailroom? My parents do, but those times are gone. The media relentlessly paints automation as the villain, but visit any country that does not enjoy massive modern automation, and it's dire poverty for everyone but the 0.01% elite. My 2cts, Cheers, Joannes
@Lokad Thank you for the answer to my badly worded question. The part I forgot about was that for those with a high school education or lower MIT, believes since 1980 70-50% of income inequality is a result of automation. They think the AI revolution will only make this worse. America has one of the worst education systems, along with the vast majority of immigrants only having a high school education or lower. How can a society function when large chucks of your population only have rudimentary education when all the new jobs require skilled labor?
I'm 3 classes away from a bachelor degree in supply chain transportation and logistic management. I work for the largest aerospace in the defense company in the world. Can a master degree in supply chain engineering or somehow becoming a supply chain scientist? Keep me safe in the supply chain domain. Thanks
Hi Kurtney, Joannes addresses this question here: czcams.com/users/live4xeV0YVRK68?si=Ug9E8lo5FmlR7dLN&t=3995. Here are the lectures he talked about: www.lokad.com/tv/tag/supply-chain-lectures/. Generally speaking, an engineering degree - solid knowledge of math, statistics, computer science combined with programming skills are and will be crucial.
@Lokad, I appreciate the information!
Thanks for sharing! Like the video and concept very much. I just wonder how to we take ordering cost into account in this framework. Suppose there are two replenishment options: Option I: restock 10 every two reorder time cycles Option II: restock 5 every one reorder time cycles Suppose both options can cover the demand. This rewards function will prefer Option II instead of Option I. But if we take the ordering (shipment...) cost into account, Option I might be the right Option. Could you please comment on that? Thanks!
Thanks for the kind word! The 'action reward' can end-up favoring either of the two options depending on the probability distribution of the demand. Indeed, if the demand is very dispersed, then, committing to Option II (bigger order) is very risky, as there is a much bigger risk of overstock; hence Option I will be favored if the inventory risk outweights the transport overhead. On the contrary, if the demand is very steady, then, the transport overhead will dominate, and assuming that the stock doesn't rapidly expire either, Option II will be favored. Hope it helps, Joannes
Wholesome and sensible conversation. Thanks for the post
Thank you! Joannes
Upfront: Really awesome lectures, it is really eye opening to see what is possible and actually useful in practice compared to classic supply chain textbook literature. I have a question regarding the broadcasting mentioned between 01:00:20 - 01:01:52. How many parameters get overall initialized in the autodiff block? (a): 1*3=3 for the <SKUs.Level> (1 for each SKU) and (b): 2*7=14 for <CD.DoW> (2 categories where each category has one parameter for each of the 7 days of the week)? And in Line 13, after having randomly "picked" a SKU, Envision "knows" which of the 14 <CD.DoW> parameters must be used for/in the Stochastic Gradient Descent because <Day> can only belong to one <DayOfTheWeek> and only one <Category> can belong to the picked <SKU>? If so, I guess I am just not used to anything being able to make/infer these connections on its own :D Best regards, Jan
Your reasoning looks correct. In the example given, we have 3 SKUs, 2 categories and 7 days a week. Thus, 'SKUs.Level' is 3 parameter values; and CD.DoW is 2 * 7 = 14 parameter values. Cheers, Joannes
Comparing the practice knowledge of Chat GPT to the common sense of a cat is completely missing the diffrence between cats common sense and instinct.
Indeed, just trying to keep the discussion vaguely relevant to supply chain challenges :-) Cheers, Joannes
Try to understand it using auto translate 😂. Lokad content always interesting for me since not a lot of people cover material management in aviation MRO
- What are these typical operation research topics that I'm saying don't work? - 11:55 - First Time series in demand forecasting - Doesn't work well for unusual demands - For example, where you have consumers who buy switches one at a time and construction companies who buy 500 at a time. - The order for 500 is announced well in advance, while the one offs are immediate, yet a time series will not differentiate between these cases and squash the demand together - For substitutions in fashion - if the wrong size is there - can't sell it, but if its a slightly different color, the customer wil likely still take it - Diapers and market basket effects - Diapers are expensive and have high brand loyalty on average - However it's not the loss of the diaper sale that is the most impactful - it's the fact that parents will then not buy all of their other groceries at the hypermarket if the diapers aren't there. - Time series also hides the fact that the future is blurry and in which direction - Deterministic forecasting works pretty well with consistent values - But a deterministic forecast of something like a soup in a supermarket with promos and demand swings of >100% it makes no sense to generate a point forecast - 1st order effects can be measured such as impact of promotion, but those indirect 2nd order effects e.g. consumer behaviour changes are hugely critical to take into account - 27:15 - For example a spare parts organization of an aircraft company had a recommendation to purchase a part, the employees said absolutely not! - It seemed a reasonable recommendation to make - however it was a spare part for a 747 - and since the part had a 30 year life, and the 747s are being deprecated within the next 10 years there was no need for it - Another interesting effect in airlines is the one-way standard. - A plane is allowed to have a part matching the old standard, however as soon as a part passing the new standard is placed - you must only use new parts going forward. - This means if you have inventory of old and new parts - each time you replace with a new part- you are modifying the composition of your demand for spare parts across your flotilla - Just because we don't know what to optimize for, doesn't mean we can't find out - 29:17 - It just can't be done in a top down cartesian way in which we split the problem down into constituent pieces and come up with an answer that way - We arrived at the need for supply chain scientists to operate under the idea of Experimental Optimization - To optimize you must create logic to generate a decision - you will then have people who object - Their reasons for objecting are typically correct - Use anecdotes to find the reasons it won't work - Then we will feedback into the dollarized/financialized decision making system to add the constraints and requirements that will meet the edge cases which are critical to that customer - For example with the spare parts problem we had taken into account the lifespan of the part, but not the lifespan of the plane itsself - In our supply chain books we tell you the demand is a Gaussian, lead times are a Gaussian - is there any way to falsify this? no, it's in the abstract, divorced from reality - 31:50 - Our goal is to make a mathematical model , maximized for modeling reality, not necessarily mathematical simplicity - An important part of the process is making sure to present the results to the customer in a specific way - not just "here are your optimal stock levels" but - "where should I put my first Euro of inventory investment?" - 80% customers know what they're doing - so this list of prioritized investments in purchasing, manufacturing, or inventory levels is reasonable - though sometimes they may be missing something obvious - Working with a german MRO company - Retrofit and repairs are 2 different things - Repairs are - engineer says this part needs to be fixed, replace - Retrofits are - the manufacturer has some doubts about a part and requires a push to replace within a month - You can't mix push and pull demands together - they act completely differently - You push a big spike of parts - and now all of your spares for that part are synchronized
Awesome summary! Cheers, Joannes
Johannes, you make a good point about the spherical cow assumption in core engineering not ending up in field calculations, but the same not being true in supply chain. Very astute observation. For someone who transitioned into “data science” from mechanical engineering and then operations research, in my limited experience, I can propose a reason as to why this is the case. In engineering, we rarely had managers who weren’t engineers themselves first. So, everyone sort of spoke the same language of math and physics and simulations and it was easier to have conversations regarding these. However in supply chain, what I’ve found is, there is, by default organisational structure, a split between the planning side (scientists versed with stats) and the implementation side (planning managers who rely on “experience”). And somehow, the implementation side calls all the shots, so more often than not, the technical solution that is chosen by the company depends on the level of technical expertise of the most technically challenged planning manager.
Astutely observed. Indeed. Cheers, Joannes
Fantastic talk! Tkank you