A Data Science Central Community
For full blog post on the 4th annual P&OMs Conference:
Hello! This week I am blogging from the 2012 Production & Operations Management (P&OM) 4th World Conference in Amsterdam. An exciting conference for analytics professionals indeed. Lest you applaud my travelling initiative, I should disclose I am currently based in the Netherlands, so my attendance at the conference was somewhat opportunistic. However, this should not at all detract from the scope and breadth of this influential conference, which is being attended by scholars from far-flung locations as far away as Brazil, Malaysia, New Zealand, and Germany
Amsterdam is an excellent historical setting for a conference on Operations Management. As per an earlier blog posting, Amsterdam during the 17th century became a nexus of global trade, thrust forward by the first genuinely modern global multi-national firm, the Dutch East Indies Company (VOC). The techniques employed by the VOC for financing, planning, and undertaking world-spanning trade expeditions would not have been possible without late-Middle Age advancements in supply chain and operations management, including advanced metrics-based management, a byword for what we today call analytics. Risk analysis, derivatives-based risk sharing, centralized record keeping combined with urban infrastructure planning, and, of course, advanced engineering and navigation methods for the associated major merchant shipping endeavors.
For those analytics practitioners who have not had an opportunity to attend a formal academic conference, I highly recommend it. It is an opportunity to gain insight into the methods and trends developing in a particular field based on advanced research, often several years before the developments potentially evolve into commercial applications. Beyond having an opportunity to attend an array of stimulating presentations based on extensive research, often involving multi-year efforts and, at times, consortia of talented scholars, the serendipitous opportunity to meet socially and talk informally with world-wide experts in a particular field can often bring sudden insights and contact with promising concepts. If you are entrepreneurial minded, attending a conference can often bring a winning business idea based on a random presentation or meeting. For instance, I met Lauri Lattila, a PhD from LUT Kouvola and we spent some time discussing his compelling practitioner research related to discrete-event based supply chain and transport logistics efficiency planning in Finland (using AnyLogic).
Before getting into the meat of the conference, I would like to briefly situate Operations Management as what can be genuinely called the ‘Godfather’ of Analytics. Operations Management, and its allied discipline, Operations Research (the two are at times claimed as the same, though I would assert the latter is much more ensconced in pure research surrounding mathematical efficiency methodologies as opposed to broader commercial application), can be considered the ‘engine’ which has propelled ‘analytics’ to its current fever-pitch state. The efficacy companies such as Dell, Wal-Mart, Amazon, and Zara (to take a EU example) have achieved has undeniably been achieved by adopting analytics-based operational efficiency as a central strategic driver. Given this demonstrated efficacy which has piqued the interest in analytics from companies in all industries, it becomes interesting to ask where the forefront of Operations Management research is today. By understanding how this field is evolving, we can gain an appreciation of where analytics is developing in the mid-term future.
Major topical trends at the conference include the following: sustainability, behavioral decision making, complex supply chains, global multi-organizational collaboration, computer-based simulation, and stakeholder analysis. One presentation I attended, entitled ‘Is Operations Management Relevant?’ (Doran, D., Hill, A., Ektas, B) got to the root of things by examining the topical schism between practitioner concerns and OM teaching. Indeed, one issue is that the increasingly advanced computational decision management techniques issuing from Operations Management research are creating somewhat of a ‘logjam’ in the typical organization. That is to say, the rapid advance of sophisticated techniques for optimizing commercial operations take time to be digested by organizations. Indications are that the academic world is a good decade in the future concerning computational sophistication in operational efficiency methods, with the exception of firms, such as the aforementioned, which have chosen to embrace OM as a strategic imperative (and are realizing outsized benefits as a result).
Behavioral Operations is a significant trend in the OM discipline from this perspective. OM has recognized that its own innovations are encountering natural limitations, the most impactful of which is the increasingly apparent fact that we humans, both as individuals and in groups, often make quite poor decisions, especially in the context of efficiency planning related to complex supply chains. The fact that we evolved to often take cognitive short-cuts, also known as decision making biases or ‘System 1′ in the research of Daniel Kahneman, means that we are often prone to bypass complicated computational problem solving to make highly emotional decisions, or ‘gut reactions’. As well, the point that many organizations are influenced by leaders who have their own self-centered interests in mind (i.e. career advancement, empire building, short-term profit) raises the omnipresent specter of ‘agency interests’ in organizations.
Especially in the context of sustainability, as witnessed by the recent US Mortgage Bubble and ensuing Global Financial Crisis, a chain of actors acting in their own self-interest can potentially unwittingly destroy significant global systems. The chain of ‘blameless self-centered greed’ between home owners, mortgage brokers, banks, investment banks, rating agencies, securitization firms and reinsurance firms each acted quite ‘rationally’ in terms of their own interests and incentives (to maximize profit). Unfortunately, the emergent macro system, involving highly complex risk sharing mechanisms such as CDOs – Collateralized Debt Obligations (bundled debt investment packages) and CDFs – Credit Default Swaps (derivatives supposedly offsetting potential default risk on slices and dices of the CDOs), lead to toxic bubble which, it can be said, ultimately led to a generational disruption in global financial stability – a ‘risk correlation event’, the aftershocks of which are still being felt and likely will be felt for several years longer.
Operations Management, at is disciplinary ‘edge’, thus is increasingly struggling with methods to manage complex systems in a world populated by individuals who were not evolved to necessarily care about the ‘big picture’. Arguably, the Global Financial Crisis was abetted thus by Regan / Thatcher Era deregulation: the misguided notion that free markets are inherently self-regulating and profit generating – best left alone to manage risk through the ‘Invisible Hand’ magic of markets. When we begin examining the emerging global challenges of climate change, sustainability, and quality of life disparitities, it becomes clear that to the degree we do not involve advanced research into these complex systems management challenges via NGOs and other regulatory bodies, we run the risk of creating the very next world compromising bubble or systemic dysfunction, although the next one could very well cause life-threatening physical destruction as sea levels swell. If we are interested in analytics, we must then become interested in ourselves, particularly how we often are very poor decision makers, especially in large groups involved in managing complex systems. There is often the aspect of a finger-pointing lack of responsibility amongst the distributed actors – everyone reaping benefit, but no one assuming responsibility for danger and risk… Luckily we have evolved the intelligence to think beyond these challenges, as witnessed by Operations Management as a formal research discipline. We just need to work on our political sensibilities in terms of our level of trust related to complex science and analytics as related to organizational governance…