Previously, Source De[Code] listeners learned how, by connecting a digital simulation to real-world environmental data, teams using digital twins can run a prototype through a plethora of potential scenarios, identify design challenges, and modify the product before ever building a physical prototype. Ben is joined by Dr. Silviu Tuca in the third and final episode of Source De[Code]'s deep dive into digital twins to contextualize their limitless potential by demonstrating how these concepts are applied to a ubiquitous aspect of daily life: our cars.
It is possible to argue that disruptive innovation in the auto industry defined each of the modern eras. Engineering breakthroughs have shaped the human experience throughout history. We tell the story of human progress through the wheel and calendar, the printing press and cotton gin, electricity and assembly lines. Henry Ford's assembly line not only democratized automobile ownership, but also transformed industries far beyond discrete manufacturing by enabling mass production.
As we enter a new era defined by the complete integration of the physical and digital worlds, the once hardware-centric auto industry must evolve to adapt to a new software-defined era. This paradigm shift in the automotive industry from mechanical and analogue components to digital electronics and intelligent machines enabling the Internet of Things is being accelerated by digital twin enabled design and test workflows. (IoT).
This episode of Source De[Code] was especially poignant for me because I am in the market for a new car. Consumers can see the digital transformation most clearly in the rapid evolution of technology in the cars we drive. Only a decade ago, standard safety and entertainment features were unheard of. Park and lane assist, adaptive cruise control, backup breaks, and a slew of other semi-autonomous features are standard on all models.
I live in the downtown core of a large, very old city in the Northeast, where narrow one-way streets and crumbling infrastructure have resulted in constant roadworks, detours, and congested urban arteries shared by vehicles, cyclists, and pedestrians. Sensor technology in new cars is already helping to improve road safety by better equipping drivers to deal with the unexpected.
Like most drivers, I hadn't given much thought to the design and testing that has gone into ensuring these gleaming new features function as intended long before I saw them in action on the dealership lot. In this episode, Ben's guest Dr. Torin gives listeners a brief overview of two methods for testing vehicle designs. The left side of the V in the first V-shaped drive testing model represents the required elements of the proposed design--features, systems, subsystems, and modules--as defined by the Original Equipment Manufacturer. (OEM).
This line accounts for both OEM-made and OEM-outsourced components. The testing and validation required to make the proposed design a reality are shown on the right.
For decades, this traditional V model has been the gold standard for automakers because it is perfectly suited for the hardware-based testing that was previously the primary consideration in drive testing. In this model, new subsystems are integrated into the vehicle, which is then driven to the test track to monitor the features under test.
As the automotive industry incorporates more software systems into vehicle design, testing methods must keep up with the software space's rapid iterations. As a result, this model is impractical for keeping up with the rapid iteration cycles of the software industry, which are increasingly defining the automotive landscape.
Automakers are adopting the disruptive DevOps-inspired design and testing model pioneered by the auto industry's new software players. As previously discussed in Source De[Code], software design and test workflows must account for complex integrations that can impact overall product functionality.
This model is similar to the figure-eight shaped race tracks where the most powerful vehicles in the auto industry compete for dominance. "A continuous feedback loop is required to achieve digital parity," says Dr. Tuca. This digital parity serves as the foundation for simulating the car and the various environments it may encounter.
Expecting The Unexpected
At the end of the day, OEMs and consumers share a common goal: safety. The primary consideration in my search for a new car is safety--my own, that of my passengers, and the safety of those around me. I live right across the street from an elementary school.
Children dash into the road as they rush towards and away from the building, chase soccer balls, or clamber into the waiting school buses and family cars that transport them between school and home. During my time at this address, I've witnessed several close calls that have chilled me to the bone. Sensor technology in the vehicles I'm considering will allow me to ‘see' more of what's around my car and react faster to the unexpected than I could ever do alone.
While no amount of technology can replace a driver's situational awareness, sensor technology and autonomous functions compensate for the brief lapses in attention that we all experience from time to time and reduce the potentially excessive cost of human error.
When placed in this context, the need for digital twins in system and vehicle testing becomes clear. Given the critical nature of these new software-enabled features, the way they are tested must reflect the messy, chaotic landscape that we call "life," which in systems testing is euphemistically referred to as "noise."
"Noise is the second term you learn in test and measurement," says Dr. Tuca. Everything is disrupted by noise." The continuous feedback loop that connects the digital prototype to the physical environment enables far more robust design testing that accounts for the "noise" of everyday life.
Design testing can constrain ambitious automotive designers who are pushing the boundaries of what is possible in the automotive space.
"You want to push the boundaries of this testing--" explained Dr. Tuca, "-- you don't want to only simulate targets; you want to simulate the guard rails, sign posts, trees, and weather conditions" to see how the vehicle behaves under any array of conditions.
In addition to the objects that drivers will encounter, tests must account for unseen forces that can interfere with the effectiveness of vehicle sensors, such as the 5G backhaul bands that lie just above and below the frequency band in which these sensors operate. Furthermore, sensors from other vehicles can interfere with one another, potentially cancelling out the enhanced safety features of modern vehicles. OEMs can use digital twins to simulate any imaginable scenario in a software environment that is so realistic that the device under test cannot tell the difference, ensuring that flaws are identified and corrected before the rubber hits the road.
If sensor design and testing breakthroughs are the accelerator, digital twin technology is the clutch that will allow OEMs to shift into high gear and propel us towards a future in which autonomous vehicles are the consumer's first and trusted choice for safe travel.
It is possible to argue that disruptive innovation in the auto industry defined each of the modern eras. Engineering breakthroughs have shaped the human experience throughout history. We tell the story of human progress through the wheel and calendar, the printing press and cotton gin, electricity and assembly lines. Henry Ford's assembly line not only democratized automobile ownership, but also transformed industries far beyond discrete manufacturing by enabling mass production.
As we enter a new era defined by the complete integration of the physical and digital worlds, the once hardware-centric auto industry must evolve to adapt to a new software-defined era. This paradigm shift in the automotive industry from mechanical and analogue components to digital electronics and intelligent machines enabling the Internet of Things is being accelerated by digital twin enabled design and test workflows. (IoT).
This episode of Source De[Code] was especially poignant for me because I am in the market for a new car. Consumers can see the digital transformation most clearly in the rapid evolution of technology in the cars we drive. Only a decade ago, standard safety and entertainment features were unheard of. Park and lane assist, adaptive cruise control, backup breaks, and a slew of other semi-autonomous features are standard on all models.
I live in the downtown core of a large, very old city in the Northeast, where narrow one-way streets and crumbling infrastructure have resulted in constant roadworks, detours, and congested urban arteries shared by vehicles, cyclists, and pedestrians. Sensor technology in new cars is already helping to improve road safety by better equipping drivers to deal with the unexpected.
Like most drivers, I hadn't given much thought to the design and testing that has gone into ensuring these gleaming new features function as intended long before I saw them in action on the dealership lot. In this episode, Ben's guest Dr. Torin gives listeners a brief overview of two methods for testing vehicle designs. The left side of the V in the first V-shaped drive testing model represents the required elements of the proposed design--features, systems, subsystems, and modules--as defined by the Original Equipment Manufacturer. (OEM).
This line accounts for both OEM-made and OEM-outsourced components. The testing and validation required to make the proposed design a reality are shown on the right.
For decades, this traditional V model has been the gold standard for automakers because it is perfectly suited for the hardware-based testing that was previously the primary consideration in drive testing. In this model, new subsystems are integrated into the vehicle, which is then driven to the test track to monitor the features under test.
As the automotive industry incorporates more software systems into vehicle design, testing methods must keep up with the software space's rapid iterations. As a result, this model is impractical for keeping up with the rapid iteration cycles of the software industry, which are increasingly defining the automotive landscape.
Automakers are adopting the disruptive DevOps-inspired design and testing model pioneered by the auto industry's new software players. As previously discussed in Source De[Code], software design and test workflows must account for complex integrations that can impact overall product functionality.
This model is similar to the figure-eight shaped race tracks where the most powerful vehicles in the auto industry compete for dominance. "A continuous feedback loop is required to achieve digital parity," says Dr. Tuca. This digital parity serves as the foundation for simulating the car and the various environments it may encounter.
Expecting The Unexpected
At the end of the day, OEMs and consumers share a common goal: safety. The primary consideration in my search for a new car is safety--my own, that of my passengers, and the safety of those around me. I live right across the street from an elementary school.
Children dash into the road as they rush towards and away from the building, chase soccer balls, or clamber into the waiting school buses and family cars that transport them between school and home. During my time at this address, I've witnessed several close calls that have chilled me to the bone. Sensor technology in the vehicles I'm considering will allow me to ‘see' more of what's around my car and react faster to the unexpected than I could ever do alone.
While no amount of technology can replace a driver's situational awareness, sensor technology and autonomous functions compensate for the brief lapses in attention that we all experience from time to time and reduce the potentially excessive cost of human error.
When placed in this context, the need for digital twins in system and vehicle testing becomes clear. Given the critical nature of these new software-enabled features, the way they are tested must reflect the messy, chaotic landscape that we call "life," which in systems testing is euphemistically referred to as "noise."
"Noise is the second term you learn in test and measurement," says Dr. Tuca. Everything is disrupted by noise." The continuous feedback loop that connects the digital prototype to the physical environment enables far more robust design testing that accounts for the "noise" of everyday life.
Design testing can constrain ambitious automotive designers who are pushing the boundaries of what is possible in the automotive space.
"You want to push the boundaries of this testing--" explained Dr. Tuca, "-- you don't want to only simulate targets; you want to simulate the guard rails, sign posts, trees, and weather conditions" to see how the vehicle behaves under any array of conditions.
In addition to the objects that drivers will encounter, tests must account for unseen forces that can interfere with the effectiveness of vehicle sensors, such as the 5G backhaul bands that lie just above and below the frequency band in which these sensors operate. Furthermore, sensors from other vehicles can interfere with one another, potentially cancelling out the enhanced safety features of modern vehicles. OEMs can use digital twins to simulate any imaginable scenario in a software environment that is so realistic that the device under test cannot tell the difference, ensuring that flaws are identified and corrected before the rubber hits the road.
If sensor design and testing breakthroughs are the accelerator, digital twin technology is the clutch that will allow OEMs to shift into high gear and propel us towards a future in which autonomous vehicles are the consumer's first and trusted choice for safe travel.