Coming to Terms with Customization
Planning a trip to China? In Beijing, don’t even think of taking a roll of toilet paper out of a public restroom. The Chinese sometimes use video camera technology; think Big Brother meets Big Data. In public restrooms in the larger cities, the Chinese have placed new facial recognition cameras which link a photo of your face to big data. Steal the toilet paper and voila– your mug could be a data-match, caught on camera!
China has moved far ahead of the West in both face and voice recognition. The Chinese language is rooted in characters, and typing translations is tricky and takes time. So it is both natural and necessary for them to leapfrog ahead in venture capital investment in many AI breakthroughs. Today, millions of Chinese simply swipe their smartphones to authorize payment.
For the rest of us, from Silicon Valley tech players, to most manufacturers and even our own government, there is a need to quickly come to terms with AI and to invest in this technology. If data can’t be pooled, the algorithms that run autonomous cars and other products won’t be efficient.
For many auto manufacturing giants, AI’s promise of customization and personalization is now a priority. While Americans still seem to want traditional non-electric cars, automotive manufacturers are choosing to bend to the forces that drive technological investment for electric and driverless cars to stay competitive. Automakers like Ford, GM, Toyota and Tesla invest billions of dollars in new smart car technology.
Tesla is only focused on smart car technology and all its cars fall into the customized model mode. Elon Musk recently claimed he is about to enter ‘manufacturing hell’ for his Model 3 electric sedan. Retailing for up to $60,000, this Tesla car faces big challenges in the customization manufacturing process. Tesla hopes to solve faulty manufacturing issues by trying to build an independent plant in Shanghai, China, which will more easily serve its largely Asian buyer. Toyota, Ford, and GM are all working on their own battery-style cars which try to offer a low cost alternative to the Tesla. Ford is involved with driverless car technology using a laser system, an astounding development for this conventional Detroit automaker.
Historically, American auto-manufacturing has been defined by assembly line-style production. This enabled large volumes of cars to be manufactured and sold with good profit margins. Manufacturers relied on these forecasts for make-to-stock models of standardized cars and trucks, and even now, most customers find these products fit their needs.
Customization however is the new trend. The make- to-custom car model is on the rise, and a challenge for most make-to-order manufacturing operations. In the customized product process, the relationship between customer and manufacturer must be seamless for deadlines and quality standards to be met. There are big challenges for manufacturers producing a more customized product. While consumers want it, they don’t want to pay more for it or wait longer for delivery. Manufacturing is challenged to eliminate waste to deliver that final product and each step along the line adds lead time and costs which need to be controlled somehow.
Make to Stock vs Make to Order
Many manufacturing enterprises rely on a make-to-stock (MTS) or build-to-stock (BTS) strategy. This build-ahead traditional production approach uses sales forecasts and historical demand data to match production and inventory with consumer demand forecasts. The MTS/BTS method requires an accurate forecast of demand in order to determine how much stock is needed for production. Good inventory planning keeps the MTS manufacturer profitable; they rely on strong forecasting from accurate historical data of sales and inventory.
All planning and purchasing concerns ultimately become production concerns, and a manufacturer’s cloud based ERP (Enterprise Resource Planning) system will quickly integrate multiple ‘demand and supply’ data forecasts, and solve material shortages before jobs go onto a production timeline. Replenishment time is a key influence in the MTS automotive operation; accurate internal lead time remains necessary to carry enough stock to cover variabilities of demand.
MTS and MTO manufacturing requires the lean automotive operation be able to transition efficiently from one form to the other. MTO manufacturing only begins with the customer order so the shop floor needs speed, flexibility and reliability to quickly produce to satisfy demand. The availability of machines to be free to do various jobs with material in stock is absolutely necessary. Transparency will aid capacity planning, fast inventory turnaround, and constant access to vendors and supply chain issues. ERP will help to manage the process of both MTS and MTO and enable the hybrid or mixed-mode shop to manage stocking constraints, lead times, and on time delivery.
Engineer to Order and Custom Order
Producing the new customized ‘smart’ car with its AI technology for self- driving, or driverless models adds completely new challenges to the shop floor. These last two methods of customization land in the ETO (Engineer to Order) and/or the Custom Order process. The smart car product requires more efficient quoting and order processing. The ability to stock to future predictions, and to deal with multiple constraints in trying to accommodate customization are the biggest challenges. The new smart car assembly is an ETO operation because engineering is required for the entire design and production process. ERP is vital to the success of ETO manufacturing since engineering is highly data dependent. Engineers need good data to move efficiently through the various stages and to be able to reduce the time needed for planning and production. Decreasing waste and reducing production time are pivotal in lean manufacturing principles, and ultimately ERP aids all ETO software operations to help engineers increase efficiency and maximize profits.
With any engineered-to-order and custom-ordered car, especially if it’s a smart car, complications in the manufacturing process can be harder to solve. For the smart car, the right data becomes crucial to the order. It is one thing to produce a car for a client who wants a red sedan with a tan leather interior, but quite another to make a customized smart car. Manufacturing technology for the smart vehicle is complex; the advancements in AI technology are dependent upon data and algorithms, and as expressed earlier, complicate efficiency. More desirable for high end American and European clients, yet more popular with the Asian mass market, this new customized smart car is the current trend and could be the future. While most auto manufacturers are ‘playing’ to markets for both traditional and customized models, Tesla is betting its future on custom-ordered smart models, since that’s the only way they make them. In China, that’s fast becoming the only way they buy them.