INNOVATIONS IN SIMULATION
For many people the year 2020 carries the aura of a particularly disruptive year.
In the auto industry, the global lockdowns in the fight against COVID-19 saw strategies that were once beneficial prove untenable. The popular notion that “leaner is better” still held true financially, but these lean systems have proved easily disrupted in a world of shutdowns, closed borders, and uncertainty.
Yet, for those in the automotive industry, the disruptions of 2020 are only the latest in a series of disruptions that are fundamentally challenging the way that we operate.
First there has been the trend towards globalization of production and supply chains. Costs were cut, savings were made, and automakers were able to produce more cars at lower rates as they took advantage of cheaper labor and transnational logistics operations. Three types of disruptions, in particular, have thrown these supply chains into peril:
Little surprise, then, that there is a move afoot to reshore the previously offshored production, and make those supply chains easier to manage.
Then there is the electrification of vehicles. The industry is being forced to both redesign and retool around a new engine, the first significant industry wide change of its sort in over a hundred years. It means new partners, new suppliers, and a brand new ecosystem for automakers worldwide. It means gambling on hybrid technology and betting on how long that hybrid reality will continue. It also means making investments in new industries and in new R&D programs as the shift towards electric vehicles moves from early mover interest to mainstream.
And then there are changing market perspectives about mobility in general. What started with ride-sharing might expand to shared vehicle ownership, or even consumers and businesses choosing to eschew vehicle ownership altogether. As cities embrace car-free zones and consumers exhibit preferences for locally sourced and zero-emission products, automakers are forced to adapt and diversify their core offer and plan for an increasingly uncertain demand for their vehicles in the future. The total industry volume (TIV) scenarios we see are now spread between -30% and +10%, a reflection of the deep uncertainty over the future direction of our industry.
Any of these disruptions would be significant by itself. For the auto industry, though, there are facing all of these plus the particular disruptions of 2020 all at once.
The tools we have today are insufficient to respond to the challenges the industry faces.
Many in the industry continue to push forward armed only with Excel and other spreadsheet tools, or invest in AI and data science to develop a clearer view of their business. Yet these tools – alone or in concert – make it almost impossible to take a holistic view of a business whose different parts are so incredibly heterogeneous.
Existing technologies in the auto industry don’t make it easy for decision makers to test more than a handful of scenarios. This limits their visibility of the entire industrial system and puts systems goals like resiliency, robustness to threats, and the capacity to anticipate and respond to disruptions out of reach. Instead of having a vision informed by testing thousands of scenarios against each other, the automaker is left to choose the best from among the few that they have, a far from optimal outcome.
And even where an auto manufacturer has embraced modern machine learning simulation technologies they are limited in the capacity to assess their systems holistically. The intricate interconnections and interdependencies are impossible to map with machine learning based-only simulation. The profitable development of autonomous vehicles beyond today’s low level is a good case study of such interdependencies, involving increasing attractiveness amongst consumers, an evolution of infrastructure supporting vehicles to grid communication (including the rollout of 5G networks), and the increasing weight of tech pure players.
As a result, significant parts of the corporate value chain are left unanalyzed and the impacts of decisions in one part of the business on other parts of that same business cannot be quantified. It’s a limited view of the value chain and it can cost a company dearly.
The end result is that the business suffers very heavy costs as it can never test more than a handful of future scenarios and can never adopt a truly holistic view, but luckily a new era of technology tools has emerged that is up to that task.
Digital twins are the key to generating accurate forecasts and testing scenarios right across the auto industry. They help decision makers test assumptions about the future state of a supply chain, the future of demand curves – indeed, assumptions about any part of the global auto value chain. They offer the capacity to test different response scenarios against each other, identifying the strengths and weaknesses of an approach before it is implemented, and allowing for the selection of an optimal course of action.
What’s more, digital twins allow decision makers to identify the points in their supply chain that are the most critical. These points can be reinforced, made more resilient and more robust so that the supply chain can absorb shocks, disruptions, and upsets that might otherwise bring it to a halt.
Prescriptive Simulation Twins are capable of modeling and simulating the entire automotive value chain and adding value across the whole auto business. It’s a technology that is both holistic and targeted, it’s dynamic and capable of responding to shifts in the industry whether at the machine, plant, process, or organizational level. Yet while the applications are endless, there are three use cases in particular where digital twin technology can unlock trapped value rapidly and with significant effect when coupled with simulation.
The first is in the new product development phase where assumptions, parts, suppliers, and logistics chains need to be compared, concentrated, mapped and charted. By adopting digital twin technology and testing different new product scenarios against each other, an optimal product development strategy can be identified that adheres to all of the constraints of the business.
The second is in the manufacturing and production process, the most capital and labor intensive part of the vehicle production chain. Digital twins with a simulation capacity have already proven effective in cutting costs, overcoming bottlenecks, optimizing production plants, and generating next-quarter returns on investment.
The third is in budgeting, an area where current tools and technologies rarely allow for the testing of more than a handful of scenarios before a final operational and strategic plan is adopted. A digital twin with the capacity to test hundreds, even thousands of different scenarios against each other, identify an optimal scenario, and generate an optimal budget offers clear advantages over spreadsheet tables and basic forecasting tools.
Digital twin technologies will be key in helping automakers navigate the current disruptions as well as the disruptions to come. While they cannot help the auto industry avoid uncertainty, they can help in building more robust and resilient systems so that local, regional, or global disruptions do not pose the sort of threats they do today. Simulation based prescriptive analytics help automakers identify and execute their optimal operational and strategic plans and prepare themselves for the dynamic and uncertain future that no-doubt awaits.
Serge Yoccoz, Cosmo Tech senior advisor and CEO of DIGITFORHUMAN, is an Automotive industry expert and digital transformation thought leader. Having held key roles at some of the world’s largest automakers, led the digital transformation team at Renault, and helped launch one of Europe’s most successful electric vehicle programs, Serge helps to drive adoption of Cosmo Tech game-changing AI-Simulation Technology.
