Executive Briefings

Preparing for Uncertainty - How 'Develop-to-Order' Companies Benefit from Flexible Development Capacity

We are in a period of unprecedented uncertainty. The pace and timing of pickup remains unclear. This poses a significant dilemma on Develop-to-Order companies in such fields as aerospace, shipbuilding and EPC. Such companies need to do significant customization to their products as the orders are processed. With the uncertainty, customer needs could change suddenly and consequently impact the order books of these companies. A single order cancellation or order addition can make the current capacity out of line with the demand. The teams could therefore be straddled with excess or incorrect resources, thus impacting the profitability. Pruning engineering capacity in tough market conditions produces another risk - not being able to meet the demand when needed. Traditionally, companies have focused on flexible capacity management for manufacturing, but this is getting increasingly important for engineering departments.

The imperative of capacity flexibility for design and development  departments is especially high in the current uncertain climate. While the pressure on controlling operational cost is intensifying, leading to shedding excess capacity, not being able to fulfill demand when needed could lead to a lost customer.

While the concepts explained below are also applicable across the value chain, including manufacturing, the focus here is primarily on the design and development domain. This is because flexibility, as a strategy, is more overlooked in design and development than in manufacturing.

Understanding the Cost of Capacity Error

Before companies build a capacity management strategy, it is important to get a view on cost of erroneous capacity. The cost comes from two kinds of errors:

• Type A Error: Demand is less than available design & development resources, leading to excess capacity

• Type B Error: Demand is more than available resources, leading to delay/cancellation of orders

This is illustrated in the figure below


Generally, the cost model is not symmetric. Type A error leads to excess cost while a Type B error is the cost of 'stockout' and is generally higher than excess capacity cost per unit. The stockout cost constitutes cost of filling a backlog of orders, costs associated with lost sales and customers. The cost models also vary from company to company, and situation to situation.

Since the demand is unpredictable, especially in uncertain economic times when there could be sudden order cancellations, there is likely to be a cost associated with wrong capacity, either from excess capacity, or from capacity shortfall. For the sake of simplicity, we have assumed a normal curve of demand deviation from capacity, with capacity planned at the most probable demand.


If we multiply the this probability with cost, we get a curve of the likely cost of demand-capacity mismatch:



Companies need to minimize the area under this curve to minimize the loss from erroneous capacity.

While a detailed mathematical approach to calculating these costs would be good, in real life enough data is not available to conduct a mathematical assessment of the cost of capacity error. Neither is the effort in building an extremely detailed model typically warranted.

However, we recommend that organizations have a strategy to minimize the cumulative losses that can be generated from mismanaged capacity. In our view, such strategy would have three elements:

1)  Design optimal capacity to minimize cumulative loss
2)  Select a combination of mechanisms to make capacity flexible
3)  Collaborate closely with engineering services partner


Planning for Optimal Capacity

The capacity that organizations should plan for should not be the same as the most likely demand. Let us assume the company keeps an incremental buffer capacity, as shown in the figure below:



This leads to a new loss curve, as shown in the next figure below. The area under the new loss curve is less than the area under the original loss curve. The overall purpose of the organizations should be to minimize this area.



Deploying Mechanisms for Capacity Flexibility

There are a number of strategies companies can adopt to introduce flexibility into their design and development to reduce their overall cost of capacity error. Many of these are learnings from capacity planning strategies applied during manufacturing, but not applied adequately in design and development. Below are a few of these strategies that are especially useful. 

Creating Capacity Options

Companies can buy some options for having capacity when needed at  short notice, without retaining the dedicated captive model. For example, if a company requires competence in structural engineering and mechanical systems, it could have some structural engineers cross-trained in mechanical systems and vice versa. Thus if the demand goes up in one of these areas, it could be met by engineers from other areas. This could be supplemented by detailed job manuals and quality gates to ensure that the work delivered by the engineers is of appropriate quality, despite their limited experience in the field. Another example would be for a company to distribute work to more than one vendor, even though it leads to loss of scale economies. The benefit is that if one of the vendors cannot meet demand, it can potentially be fulfilled by another.

Yet another example would be to give a combination of guaranteed and additional flexible workload to suppliers. This would transfer a part of the risk to the vendors, but the cost of such risk transfer would be minimal. Creating such options requires investment, but companies can estimate the cost of such options and decide which of these strategies works best for them.

Rolling Forecast Mechanism

Just like manufacturing resource planning, companies can do engineering resource planning, which starts with a periodic rolling forecast. A part of the rolling forecast is considered 'firm' and the subsequent one is considered 'tentative.'  The demand forecast triggers of a number of activities, e.g., resource confirmation, training, etc., which allows capacity to be built/released in anticipation of the demand.

Layered sourcing

While applicable in all scenarios, this strategy is especially suitable when the cost of unmet demand is huge. Companies can build a layered sourcing strategy with a core captive capacity that will take care of the base load. Any minor fluctuations can then be taken care of by the second capacity reservoir that can be called upon on short notice. Further increase in demand can then be fulfilled by the third reservoir that is slightly more expensive. Companies should list all their capacity reservoirs and analyze them along the dimensions of unit cost and lead-time. They can then assess which reservoir they could tap into as the demand for D&D team flexes.

To ensure that they are not in for any unwanted surprises, it is also essential to ensure that the assessment of these reservoirs is maintained up to date and, if possible, tested on a periodic basis.

One of the disadvantages of nurturing multiple capacity reservoirs is some loss of scale economies.  There are however, some approaches where this could be mitigated, such as organizing such reservoirs within a single vendor. 

Case Example

A large European aerospace manufacturer has been working with a well known IT and engineering services vendor for approximately a decade. They signed a strategic deal for a large design work package to be delivered by the vendor. The volume of these design activities is variable and therefore needs an organizational model that allows it to scale up (or down) when needed at a short notice.

The two companies worked together to create a model in which the vendor has a core team of people working for the work packages. This is supplemented by a set of resources that are trained for the manufacturer's context; however, these people will be working for other clients of the vendor. At any time, when required, some of these resources can be called upon if the demand exceeds capacity. In the unlikely event of the demand exceeding even this capacity, the vendor would use its wider engineering resource pool that could be trained and put on the work. For this, the vendor has set up a client specific training academy that can train these resources at a very short notice.

This model will be supplemented by a 6-month rolling forecast that allows various actions to be triggered to prepare for the future demand. The manufacturer and the vendor meet on a monthly basis to form a view of likely demand so that the vendor can ensure adequate resources and the necessary tools for meeting the demand.

Need for improved Engineering Vendor Collaboration

Building such flexible organizations needs companies to work in close collaboration with their vendors. This is becoming increasingly important as the business models of such companies evolve where outsourcing design and development is getting increasingly acceptable, yet the organizations would like to retain strategic agility. Such models therefore require a partnership-based approach rather than transactional approaches that keep arm's-length distance from the vendors.

Benefits

The benefits of applying a more scientific approach to design and development capacity can be great.  In develop-to-order companies, design and development constitutes a significant part of the overall spend. Mismatch in demand and capacity could potentially wipe out a significant part of the profits of the company. While the authors have not seen a scientific study in this area, their experience indicates that a scientific approach to capacity management could save as much as 5 percent to 10 percent of the design and development cost for such companies. 

Authors: 

Mark Robinson is the head of Project and Programme Management, Integrated Body of Knowledge, at Airbus SAS in Toulouse, France. 

Pankaj Chugh, responsible for market strategy for the manufacturing segment of Infosys Technologies, leads his company's relationship with Airbus. 

Source: Airbus & Infosys Technologies

We are in a period of unprecedented uncertainty. The pace and timing of pickup remains unclear. This poses a significant dilemma on Develop-to-Order companies in such fields as aerospace, shipbuilding and EPC. Such companies need to do significant customization to their products as the orders are processed. With the uncertainty, customer needs could change suddenly and consequently impact the order books of these companies. A single order cancellation or order addition can make the current capacity out of line with the demand. The teams could therefore be straddled with excess or incorrect resources, thus impacting the profitability. Pruning engineering capacity in tough market conditions produces another risk - not being able to meet the demand when needed. Traditionally, companies have focused on flexible capacity management for manufacturing, but this is getting increasingly important for engineering departments.

The imperative of capacity flexibility for design and development  departments is especially high in the current uncertain climate. While the pressure on controlling operational cost is intensifying, leading to shedding excess capacity, not being able to fulfill demand when needed could lead to a lost customer.

While the concepts explained below are also applicable across the value chain, including manufacturing, the focus here is primarily on the design and development domain. This is because flexibility, as a strategy, is more overlooked in design and development than in manufacturing.

Understanding the Cost of Capacity Error

Before companies build a capacity management strategy, it is important to get a view on cost of erroneous capacity. The cost comes from two kinds of errors:

• Type A Error: Demand is less than available design & development resources, leading to excess capacity

• Type B Error: Demand is more than available resources, leading to delay/cancellation of orders

This is illustrated in the figure below


Generally, the cost model is not symmetric. Type A error leads to excess cost while a Type B error is the cost of 'stockout' and is generally higher than excess capacity cost per unit. The stockout cost constitutes cost of filling a backlog of orders, costs associated with lost sales and customers. The cost models also vary from company to company, and situation to situation.

Since the demand is unpredictable, especially in uncertain economic times when there could be sudden order cancellations, there is likely to be a cost associated with wrong capacity, either from excess capacity, or from capacity shortfall. For the sake of simplicity, we have assumed a normal curve of demand deviation from capacity, with capacity planned at the most probable demand.


If we multiply the this probability with cost, we get a curve of the likely cost of demand-capacity mismatch:



Companies need to minimize the area under this curve to minimize the loss from erroneous capacity.

While a detailed mathematical approach to calculating these costs would be good, in real life enough data is not available to conduct a mathematical assessment of the cost of capacity error. Neither is the effort in building an extremely detailed model typically warranted.

However, we recommend that organizations have a strategy to minimize the cumulative losses that can be generated from mismanaged capacity. In our view, such strategy would have three elements:

1)  Design optimal capacity to minimize cumulative loss
2)  Select a combination of mechanisms to make capacity flexible
3)  Collaborate closely with engineering services partner


Planning for Optimal Capacity

The capacity that organizations should plan for should not be the same as the most likely demand. Let us assume the company keeps an incremental buffer capacity, as shown in the figure below:



This leads to a new loss curve, as shown in the next figure below. The area under the new loss curve is less than the area under the original loss curve. The overall purpose of the organizations should be to minimize this area.



Deploying Mechanisms for Capacity Flexibility

There are a number of strategies companies can adopt to introduce flexibility into their design and development to reduce their overall cost of capacity error. Many of these are learnings from capacity planning strategies applied during manufacturing, but not applied adequately in design and development. Below are a few of these strategies that are especially useful. 

Creating Capacity Options

Companies can buy some options for having capacity when needed at  short notice, without retaining the dedicated captive model. For example, if a company requires competence in structural engineering and mechanical systems, it could have some structural engineers cross-trained in mechanical systems and vice versa. Thus if the demand goes up in one of these areas, it could be met by engineers from other areas. This could be supplemented by detailed job manuals and quality gates to ensure that the work delivered by the engineers is of appropriate quality, despite their limited experience in the field. Another example would be for a company to distribute work to more than one vendor, even though it leads to loss of scale economies. The benefit is that if one of the vendors cannot meet demand, it can potentially be fulfilled by another.

Yet another example would be to give a combination of guaranteed and additional flexible workload to suppliers. This would transfer a part of the risk to the vendors, but the cost of such risk transfer would be minimal. Creating such options requires investment, but companies can estimate the cost of such options and decide which of these strategies works best for them.

Rolling Forecast Mechanism

Just like manufacturing resource planning, companies can do engineering resource planning, which starts with a periodic rolling forecast. A part of the rolling forecast is considered 'firm' and the subsequent one is considered 'tentative.'  The demand forecast triggers of a number of activities, e.g., resource confirmation, training, etc., which allows capacity to be built/released in anticipation of the demand.

Layered sourcing

While applicable in all scenarios, this strategy is especially suitable when the cost of unmet demand is huge. Companies can build a layered sourcing strategy with a core captive capacity that will take care of the base load. Any minor fluctuations can then be taken care of by the second capacity reservoir that can be called upon on short notice. Further increase in demand can then be fulfilled by the third reservoir that is slightly more expensive. Companies should list all their capacity reservoirs and analyze them along the dimensions of unit cost and lead-time. They can then assess which reservoir they could tap into as the demand for D&D team flexes.

To ensure that they are not in for any unwanted surprises, it is also essential to ensure that the assessment of these reservoirs is maintained up to date and, if possible, tested on a periodic basis.

One of the disadvantages of nurturing multiple capacity reservoirs is some loss of scale economies.  There are however, some approaches where this could be mitigated, such as organizing such reservoirs within a single vendor. 

Case Example

A large European aerospace manufacturer has been working with a well known IT and engineering services vendor for approximately a decade. They signed a strategic deal for a large design work package to be delivered by the vendor. The volume of these design activities is variable and therefore needs an organizational model that allows it to scale up (or down) when needed at a short notice.

The two companies worked together to create a model in which the vendor has a core team of people working for the work packages. This is supplemented by a set of resources that are trained for the manufacturer's context; however, these people will be working for other clients of the vendor. At any time, when required, some of these resources can be called upon if the demand exceeds capacity. In the unlikely event of the demand exceeding even this capacity, the vendor would use its wider engineering resource pool that could be trained and put on the work. For this, the vendor has set up a client specific training academy that can train these resources at a very short notice.

This model will be supplemented by a 6-month rolling forecast that allows various actions to be triggered to prepare for the future demand. The manufacturer and the vendor meet on a monthly basis to form a view of likely demand so that the vendor can ensure adequate resources and the necessary tools for meeting the demand.

Need for improved Engineering Vendor Collaboration

Building such flexible organizations needs companies to work in close collaboration with their vendors. This is becoming increasingly important as the business models of such companies evolve where outsourcing design and development is getting increasingly acceptable, yet the organizations would like to retain strategic agility. Such models therefore require a partnership-based approach rather than transactional approaches that keep arm's-length distance from the vendors.

Benefits

The benefits of applying a more scientific approach to design and development capacity can be great.  In develop-to-order companies, design and development constitutes a significant part of the overall spend. Mismatch in demand and capacity could potentially wipe out a significant part of the profits of the company. While the authors have not seen a scientific study in this area, their experience indicates that a scientific approach to capacity management could save as much as 5 percent to 10 percent of the design and development cost for such companies. 

Authors: 

Mark Robinson is the head of Project and Programme Management, Integrated Body of Knowledge, at Airbus SAS in Toulouse, France. 

Pankaj Chugh, responsible for market strategy for the manufacturing segment of Infosys Technologies, leads his company's relationship with Airbus. 

Source: Airbus & Infosys Technologies