To Integrate, or not Integrate?
I had the pleasure to work with hardware and component development in some different areas of the industry from forklifts, Heavy Duty trucks to passenger cars and automotive during my career.
One common question and strategy that has reoccurred in all fields and areas has been the question of integration of components. Some typical examples are Electrical machines and inverters, DC/DC and Onboard chargers, and combination controllers controlling two or more actuators.
So what is the best strategy here and what are the key parameters affecting the optimal solution then?
Integration
Advantages:
• Technically possibilities to reduce weight and cost and complexity. (Use the same housing, avoid cables, cooling pipes, etc.)
• Synergies are possible, and some components can be reduced or removed if for example two ECU units can be combined possibly a lot, or all electronics PCB, uC, etc. can be combined into one.
• Less complex final assembly fewer cables, cooling pipes, etc.
One must take a strategic view and look upon the components to be integrated, make a Pugh matrix or something similar, define the different
parameters, and rate their importance for my application and my business
Disadvantages:
• If different suppliers are used this can be a challenge in many ways. Responsibilities and interfaces need to be carefully defined.
• In general, less flexible if the components are used in different vehicle models and applications.
• Thermal aspects can be a challenge if combining components with different thermal capabilities inverter/ electronics motors transmissions.
• Aftermarket, serviceability needs to be considered.
“Stand alone”
Advantages:
• Normally more available off-the-self components are available.
• Flexibility to move and pack the components to adopt different installations with the same components.
• Flexibility also for upgrades and changes on the components. Changing or upgrading of a component can be done more framed to the component than in an integrated design. Disadvantages:
• As an inverse to the integration, additional cables, harnesses, and cooling pipes are needed.
• Challenges like EMC are normally more severe in this setup.
• May lead to complex solutions in the surrounding components.
As one can see there are many parameters to consider, and these are just some examples, many more will come depending on which components are to be integrated.
Looking at also the not technical aspects some of the key parameters to consider are:
• How flexible will your system need to be? Generally, if high flexibility is needed stand alone is preferred. What is the long-time strategy?
• What will your volumes be? Will your volume motivate the customization needed for an integrated solution? Integration normally comes with higher tooling and more investments but enables lower part prices and saving in assembly and production.
• One last aspect that is worth mentioning is if you are intending to make an in-house design and production or if you intend to buy components from tier 1 suppliers. Integration makes much more sense in an in-house design where the synergies of the combination of the components can be utilized fully.
So, to wrap up let’s come back to the initial question should you integrate or not? What is the recipe for the correct decision? Well to state one of my old engineering friends’ favorite statements “It Depends”.
One must take a strategic view and look at the components to be integrated. Make a Pugh matrix or something similar and define the different parameters and rate their importance for my application and my business.
From my experience, the same question on integration has ended up in different “best” solutions in different companies/ businesses.
