Saturday, March 23, 2013

Instant Gratification Culture Demands a New Manufacturing Paradigm

The information technology revolution is creating a culture of instant gratification. People have an urge and expectation to immediately get things they want. Fortunately, meeting this requirement in the digital world is quite straightforward. A hit song, a blockbuster ebook, a viral video, or popular news can be instantly accessed by millions of people at the click of a mouse. The scaling up process needed to meet huge overnight surge in the demand can be accomplished in a matter of hours without any significant cost repercussions.

Consumers are beginning to exhibit a very short attention span as a side effect of this culture. Something that is hot today is not likely to be remembered in few weeks from now. So it makes sense for companies to meet the market demand with an utmost urgency. Making customers wait for your product is simply not a viable option because they might move on to the next “in” thing.

This culture of instant gratification is also carrying over to physical products. People want to be able to purchase a hot new product immediately, but they want to pay a bargain price for it! Unfortunately, scaling up production in the physical world is a lot different from making digital copies. So the instant gratification culture poses new challenges for manufacturers. In the last few years, several companies had difficulties in meeting sudden surges in demand. Notable examples include Apple iPhone, Toyota Prius, and Nintendo Wii.

When a new product is launched, it is often difficult to accurately estimate its demand. So production typically begins at low volumes and production capacity is gradually added to meet the growing demand. Unfortunately, this model does not work well in situations when there is a sharp increase in the demand overnight. The culture of instant gratification demands a manufacturing system with the following characteristics:
  • ability to dramatically increase production volume on a very short notice due to a sharp increase in demand,
  • short production run spans (e.g., a few months) to quickly phase out products as customers move on to the next best-seller,
  • and low production costs to ensure affordability.
We refer to this type of manufacturing as burst production, i.e., very high volume production over a short period of time.

So far, the mass production has been the dominant paradigm for producing affordable products. It is based on the economy of scale. It usually takes many months to setup a production line for the mass production. Processes are carefully optimized on the production line to reduce the cycle time and reduce production cost. It is not unusual for a production run to continue for years with minor tweaks in the product and the process. Long production runs enable amortization of the initial setup costs. This paradigm has been successfully used to manufacture affordable products and raise the standards of living worldwide.

The characteristics described above clearly show that the burst production manufacturing is a fundamentally different paradigm from the mass production manufacturing. One simply cannot take six months to setup a production line. Currently the only viable method for supporting a burst production operation is to deploy a vast amount of manual labor over a short period of time, exploiting the flexibility of the human labor to reduce the changeover time. Currently China seems to be the only place in the world where a vast amount of labor can be rapidly mobilized at a relatively low wage rate. As we try to bring manufacturing jobs back to the US, we will also need to develop a new manufacturing paradigm to support the burst production. This new paradigm will need to be supported by a number of new technological advances.

The rapid scale up requires quick access to additional equipment. Ordering the new equipment and bringing it on-line is simply not feasible because of time constraints. There are two potential solutions on the horizon. The first solution involves customers being able to 3-D print products at their homes or shops that offer 3-D printing services (e.g., Kinkos for 3-D printing). The second scenario involves companies buying additional manufacturing services from the third party manufacturing service providers. In both of these scenarios, companies will just send digital product models to customers or service providers. Realizing this vision in practice will require dramatic improvements in digital fabrication and new standards for exchanging digital product models. Ensuring manufacturing quality in the above scenarios will be a challenge. So we will need new low-cost high-throughput inspection systems to ensure that products can be made as per specifications in the digital models.

Most high-rate production processes currently use custom tooling (e.g., molds, dies). Ordering and getting these tools often takes significant time. So we will need to either develop high-rate production processes that do not require tooling, or come up with methods to quickly produce custom tools. Several processes such as 3-D printing and water-jet cutting exist that do not require part-specific tooling. However, 3-D printing is currently quite slow and water-jet cutting is mainly used for making parts with 2-D shapes. So new high-rate processes will be needed that can produce 3-D parts without requiring part specific tooling. 3-D printing is already being used to rapidly produce custom tools (e.g., direct metal laser sintering process by EOS). However, currently material options are limited. So we will need to improve 3-D printing to rapidly realize custom tools with the desired material properties.

Most manufacturing processes require complex setup and calibration processes that consume significant amount of time. Robots can play a significant role in speeding up the setup processes. Many new capabilities have been added to robot manipulators over the last few years (e.g., Barrett Arm and Hand). However, significant advances will be needed in robot manipulators to match human dexterity, flexibility, and speed. Better user interfaces and diagnostic functions can also reduce the time needed to setup machines. Many processes require manual recalibration to ensure product quality. Low-cost sensor networks are enabling the possibility of creating machines that can calibrate themselves.

Most modern factories use programmable machines. Programming machines and robots often takes significant effort and expertise. We need to dramatically reduce the time needed to program machines and robots. Programming by demonstration appears to a very promising method for reducing the programing time. Baxter, a new robot from Rethink robotics is a good example of the progress being made on this front. We will need to make significant additional progress in the programming by demonstration field to ensure that this approach is applicable to a broad class of manufacturing, assembly, and material handling processes.

Efficient logistics is an integral part of high throughput manufacturing operations. We will need logistics technology that can be quickly reconfigured and scaled up or down based on the demand. Kiva system has developed technology to provide flexible logistics support for warehouses. This technology appears to be easy to deploy and configure. We need technology with similar capabilities for manufacturing shop floors. Manufacturing shop floors tend to be much more complex than warehouses. Significant new developments will be needed in the logistics technology to support burst production.

Many manufacturing challenges can be eliminated by carefully designing products. We will need to develop the design for burst production methodology to ensure that the sudden dramatic increase in demands can be easily handled. This will require carefully selecting processes and materials to ensure that they have low lead times. Using a large number of standard and common components would be a step in the right direction. Some lessons from the design for mass customization methodology will be applicable to this area and will need to be carefully adapted to meet these new challenges.

Burst production paradigm goes well beyond the mass marketed products. For example, vaccine manufacturing can also benefit from this paradigm. A disease outbreak may generate a large demand for a vaccine that needs to produced and distributed rapidly.

The digital world has created a new breed of customers with an expectation of instant demand fulfillment. The manufacturing world has to come up with a solution to meet this expectation.