In an insightful interview, Professor Milind Sohoni, Professor, Operations Management, ISB, talks to Ananth Rangarajan, PGP Co. 16, and Jaiwardhan Gupta, PGPMAX Co. 18, about the critical challenges and opportunities for supply chain systems in India and the impact of automation, artificial intelligence and GST on supply chains now and in the future.

Key challenges for supply chain systems to gain scale and reliability

Ananth: What do you think are the key challenges for supply chain systems to gain scale and reliability in todays India?

Prof. Sohoni: Unfortunately, with supply chain management structures, there doesn’t exist a “one size that fits all.”   So gaining scale and building reliability/resilience will depend on the product/service being provided, supply chain strategy adopted and how the supply drivers have been designed.  For example, if one were to consider fast moving consumer goods (FMCG) goods, typically achieving scale would require strategies to consolidate on the demand side, i.e., gain larger market share. This in turn, will then necessitate designing supply chain levers like where and how your facilities are located (networks), how inventory management is done (including sourcing and procurement), how is your distribution/procurement network setup (including transportation and last-mile delivery issues), and what kind of information is collected to improve decision making and supply chain efficiency.  There are several other related supply chain levers that may need tweaking.  For example, if the firm wants to provide customization and variety at scale, it’ll have to worry about  how to trade-off between which products (and how much) should be made-to-stock and which products (and how much) should be made-to-order, i.e., where should the push-pull boundary exist within the supply chain? What is the form of the product that you are going to carry, i.e., how much of it can you delay producing the finished product, and  up to what point in the supply chain? Depending on the products and the processes that you have in place, you can achieve the kind of scale that you require, but it’s not a one-size-fits-all solution. One has to think about:

  • How should the firm consolidate its markets?
  • How do I design the supply chain levers (products, services, and processes) to enable delivering to scale?  For example, this could involve making choices between holding finished goods inventory and transporting on demand to achieve responsiveness in time.

To your second point about reliability/resilience there are a few things that firms can consider.  Again, let me give you an example from the procurement side. One typical way firms address reliability/resilience is by building redundancy into the system through multiple (or at least dual) sourcing options. Additionally, firms may want to build more visibility into their supply chains by using technology and collecting use of data.  Using state-of the art techniques, firms may be able to predict failures prior to the event and devise early interventions to recover in the best possible way, not necessarily in the least cost way, but in the most responsive way.  Here information gathering (big data and IoT), analytics, and the ability to recognize failure patterns (through data and information) is key. The better the information and the closer to real time it is, the better your ability to process that information is going to be. That will make a big difference.

Scaling and Strategic Positioning

Ananth: If you are scaling up, what should be your strategic positioning? 

Prof. Sohoni: A firm’s strategic positioning depends on the product/service it offers, the competition its faces, and the value proposition it emphasizes to its customer base.  Scaling up, is about providing the same set of products and services, but to a larger base.  So for example, if you provide customized products then that’s what you want the scale. Now, if by scaling up the firm’s strategic positioning were to change, then being aware of the new market, competition, and value proposition to the customer base becomes necessary for the firm to consolidate its position.  For example, the firm might want to get to ways of demand consolidation because that is what drives all the volumes through its supply chain.

Weaving Digital thread through Supply Chain
Jaiwardhan: What should companies be doing to weave a digital thread through their supply chains for better visibility of information and faster, higher-quality decision making? How can they use it to operate the supply chain more effectively?

Technology and digital supply chains are the future. Gone are the days when you could hide behind the fact that information was not readily available and that, as a result, you could not use it in planning. Today there is enough technology available both in terms of the supply and demand side for you to start doing a better job. For example, when I say technology, I am not just talking about information technology. It is also the use of data and information to do much better planning and to start thinking about improving the scale of operations.

Let me give a slightly different example to emphasize what the advent of the internet of things (IoT), digital technologies, and digitization is doing to supply chain strategies, decision making, and designing supply chain levers.  For example, consider a high-tech product like an aircraft engine (jet engine.)  These aircraft engines are very expensive and there are only a few original equipment manufacturers (OEMs) that manufacture them.   Since the 1950s, improvements in technology have significantly improved the life of aircraft engines.    The cumulative effect of advances in engine technology meant that by the 1990s when the first of the current generation of engines like the Rolls-Royce Trent entered service they not only exhibited a level of performance in terms of thrust and fuel efficiency that was a very  substantial improvement on 30 years earlier, but their durability was also vastly improved compared to first generation turbofans like the Rolls-Royce RCo42 Conway. The extent of this improvement was reflected in the demand for spares to replace worn and damaged components.

Even in the early 1980s, a jet engine would consume a quantity of spares equivalent to the original value of the engine in about eight years, but engines like the new Rolls-Royce Trent that entered service at the end of the 1990s consumed this quantity of spares in 25 years. This represented a very big drop in demand for spares and therefore their revenues.

Now think of how these supply chains morphed from being “product” supply chains to “product-as-a-service” supply chains.  Essentially, given the saturation in the product and spares market, the OEMs (original equipment manufacturers) had to start thinking about how they are going to make their supply chain sustainable and scale up.  So rather than simply focusing on being engine manufacturers, the OEMs decided to get into after-sale contract services too, i.e., provide superior, cost effective maintenance and support services throughout the after-sales phase of the customer-supplier relationship.  Essentially, the strategy is to improve the “up-time” for airlines and reduce the cost of ownership by tying their compensation to the “output” value of the product generated by the airline.  The OEMs realized that  that rather than selling engines to the airlines they would write contracts wherein the airline leased the engine and paid the OEM for keeping their engines functional, i.e., reduced downtime.  So the supply chain strategy shifted focus to re-designing the contractual and implicit relationships between airlines and OEMs in the service support supply chain.

This practice has caught on in other supply chains too.  Customers and suppliers of mission-critical products, such as semi-conductor manufacturing equipment, high-speed trains, earth-moving equipment, commercial aircraft and military weapon systems, are recognizing that the acquisition of world-class products is not sufficient, but rather it is necessary to provide superior, cost effective maintenance and support services throughout the after-sales phase of the customer-supplier relationship.  This change is sometimes referred to as “servicitization” of a product supply chain.

IoT, Big Data, and Analytics play an important role here. The examples cited, to differing extents, incorporate the use of networked sensors to collect data in order to inform management decisions. This is achieved by using sensors connected to the IoT, on assets from aircraft engines, trucks and trains in order to collect data. The sensors gather data on the health of components, potential maintenance issues, and other relevant performance indicators. This data can then be used to inform decisions that drive efficiency, cost-reductions, and other benefits for clients.

These examples also link back to the earlier question about reliability. what GE and Rolls Royce are saying is that the system reliability is high because rather than maintaining after a  failure has occurred, they are reducing downtime  because of unplanned maintenance/ failure (thus increasing revenue opportunities for the airlines.)  With today’s technology and analytical capabilities, they can predict when a particular part needs to be replaced and be proactive about replacing it.   The added benefit for them is that they can smoothen their own spare parts inventory management.  By moving towards servicitization, they can track information, have great visibility into it, and as a result, serve their clients much better. Today, almost 60% of all revenues for some of these engine manufacturers (GE, Rolls Royce) comes from these type of pay for performance contracts – popularly referred to as Power By the Hour (PBH) or Performance Based Logistics (PBL) contracts.

Today, many of us are talk about industry 4.0 and internet of things (IOT) applications, where a large amount of data is collected on a regular basis.  This information and insights gleaned of this data is then used in planning functions to improve revenues. I think that’s the future — connecting the physical world to the digital world and using data-driven insights, patterns, and inferences to improve supply chain operations.  

Today you see business models evolving that are purely because of availability of digital technology and information.  In that sense, the point you raised is very critical. We talk of shared economy models or platforms like Airbnb, OYO, Ola, and Uber – but, in my opinion, these are all examples of matching supply and demand in a digital world.

Shared capacities in competitive business world
Ananth: Would it be a smart choice to explore shared capacities (across competing players in the industry) in key capex intensive areas, i.e., warehouses/ line haul transport infrastructure/ manpower/  delivery infrastructure? Is there value to be unlocked there or would it be foolish to think of this in the competitive business world?

Prof. Sohoni:  With the sharing of capacities, firms have to be careful as to when, and where, it might work, when it doesn’t, i.e., reduces the firm’s profitability.   The idea of “coopetition” is not new – it has benefited several firms by reducing their costs.  There are several examples of alliances, partnerships between competing firms – for example, code-share agreements between airlines, sharing logistics and transportation facilities by FMCG firms – which have benefited these firms.  However, when sharing upstream capacity hurts the firm by increasing competition in the market, then firms may become cautious.  There are a few examples from the electronics industry where shared suppliers (at least for standard electronic components) is common amongst competing firms. However, sometimes, these firms do not share common suppliers (write exclusive contracts with suppliers) for some of their critical components to prevent “information leakage” about their products. 

Influx of automation in supply chain
Ananth: How do you see the influx of automation in supply chain change the equation?  Can automation be a silver bullet to tackle all sub-optimal processes in a supply chain? Or should one be careful with automation and engage in it only when the supply chain is fundamentally balanced, that is, when one’s  house is already in order?

Prof. Sohoni: You have to think about any change, or any automation, in view of what are you  trying to achieve with that automation. For example, automating rote tasks seems like an obvious thing to do.  On the other hand, automating complex tasks in a manufacturing environment might be necessary to maintain high-quality and reliability.  Of course, you can part automate too, i.e., decide where, and at what point, should a process be automated and what kind of automation should be used. But on a more philosophical note, today, with the advent of advanced technology (for example, in the manufacturing sector) and digitization, I think automation is inevitable – particularly to increase productivity, quality, and reliability.  We are going to see much more use of machines and what we loosely term as machine intelligence or machine learning (ML). To maintain quality, to maintain standardisation, and to maintain reliability and resilience, and even to be responsive and changeover quickly, firms are going to move towards more automation. And when I say automation, I don’t just mean machines doing the work, but also collecting data and moving towards data-driven optimisation. You will see more of that, I think.

AI's fit in a supply chain system
Ananth: Today, in the supply chain, our biggest enemy is the component of uncertainty: uncertainty in demand, uncertainty in supply and uncertainty related to our capacities (e.g. manpower). How do you see AI fit in a supply chain system in the near future? What key aspects of uncertainty can we confidently hand over to AI?

Prof. Sohoni: I don’t think uncertainty is the enemy – it can be an opportunity too.  On a more philosophical note, a world without uncertainty would be boring.  The real question you are asking is how do mangers deal with uncertainty within the supply chain?  I think there are several ways in which supply chain managers deal with uncertainty.  Better data/information gathering, better understanding of the sources of uncertainty, better mitigation strategies like in inventory risk-pooling, building redundancy in suppliers, robust scheduling, and other such strategies.

I think the big change that we are also witnessing is the accessibility of tools like AI and ML becoming available for operational decision making.  With IoT, availability of large and voluminous data, ability to analyse such data and harness patterns in the data/information, ability to transfer/store data/information that is easily transferable and accessible (cloud storage), and ability to analyse data in large amounts (cloud computing) managing uncertainty might become easier.  The big change is in the way decisions will be made – data will not be an “after thought” to verify if the decision model built (using the decision makers understanding of the trade-offs faced in the business) is good but will become a “necessity” to build the decision model itself (by recognizing emerging patterns and incorporating data-driven insights.)  One significant advantage is that decision making in such an environment can be tailor-made for the specific instance, as opposed to building decision models that may work well on an “average.”   Personally, I think this is a very fundamental shift in the way people are starting to think about data-driven decision making. The data are showing you what to do, rather than you deciding what to do and verifying ex-post if the decision you took was right.  Uncertainty exists. I need instant specific solutions, because I am reacting to a particular instance of that uncertainty. And how I react to a situation is where AI might help.

Supply chain for win-win situation in aviation industry
The Aviation industry has certain parts that are highly specialised in design/manufacturing requirements and hence require significant investments in both R&D and Plant & Equipment (P&E). Not having adequate parts at the right time could have significant impact to the ability of the industry in meeting its commitments to airlines. What supply chain approaches could the industry apply that would be win-win for everyone in the industry?

Prof. Sohoni:  Personally, I think the core idea of “product-as-a-service” and evolving technologies like Additive Manufacturing (AM) -- in a digital world -- might partly address this issue.  Let me try and briefly elaborate.  Typically, one of the key trade-offs in supply chains for high-tech, high-cost products exists between inventory management and expediting transportation.  For example, to reduce inventory holding costs a firm may want to centralise inventory of costly, slow-moving parts.  In fact, sometimes it may not even want to carry the finished product (but only assemble on demand).  However, when demand materializes, it may need to expedite shipment to reduce down-time for the client airline.  Such as strategy may result in significant loss of revenues for the client airline due to longer down-time on the revenue generating asset (the aircraft.)

Moving to a “product-as-a service” necessitates the supplier to minimize down-time for the airline.  So rather than waiting for demand to be generated, which then triggers the part procurement process, it might be better to proactively determine when a part needs to be repaired/replaced and plan the procurement well in advance.  This requires but better data acquisition, analytics, and forecasting capability – in addition to inventory and transportation planning.  Such contracts are now becoming prevalent in many high-tech industries – the PBH contract we discussed earlier.

The second point regarding technologies like AM.  While the technology is still nascent, GE has had some success in producing aircraft engine parts through AM.  AM allows for extreme decentralization with manufacturing very close to the client site.   This reverses the risk-pooling strategy adopted typically and allows for demand-driven production.  If this technology matures, then the possibilities seems very exciting for such industries.

GST implementation and how it affects supply chain in India
Jaiwardhan:  What are your views on the implementation of GST by the Indian government? What are the opportunities going forward for the industry and how this has affected the supply chain in India?

Prof. Sohoni: I have not studied the GST, in its implementation phase, in great detail. However, at a philosophical level, we have moved away from a multi-taxation system where the consumer surplus was much lower to a more unified taxation system which allows the unification of markets. So India becomes one market as opposed to states in India being different markets. I think that is a fundamental change and it is a welcome change. Initial indicators from several sectors, particularly the logistics sector, are that there is a significant reduction in the cost of delivery and in the cost of production itself. Since there is no redundant taxation, competitiveness is only increasing. Now, have all sectors been uniformly affected? Probably not.  Some may have worsened because of differential GST.  That said, the logistics market is witnessing significant consolidation post GST.  I think that opens up a whole new arena for improvement in the logistics market – even opening possibilities of new business models (as discussed earlier.)

Advice to young MBA graduates
Jaiwardhan: One final question: What is your advice to young MBA graduates about careers in supply chain?

Prof. Sohoni: It is a great career to be in. This is the oldest profession — matching supply and demand — and it is not going away. It will change its form and evolve, but I think it is very fundamental to any economy. I think the opportunities are immense. The kind of things that are happening today — the use of analytics, the use of AI, and the kind of technologies that are now getting deployed in terms of product manufacturing and in terms of services generation — these are interesting.  Remember, folks working in this domain create products, create processes, and are core to any business operations. I can’t underemphasise this fact. To all MBA graduates who are thinking about a career in this area, I can say, you will learn something very fundamental. You may not be the highest paid person in the company, but that’s fine because the job that you will do will be very satisfying.