Cognitive Business: AI and the Autonomous Driving Ecosystem
Thanks to a mutual friend at Singularity University, in this Cognitive Business interview, we explore the world of Artificial Intelligence (AI) and the Autonomous Driving Ecosystem with Justyna Zander, Software Architect and Autonomous Driving Ecosystem Lead at Intel.
Not having come across many women in the field, I had to ask: what is it like to be a woman in tech rockstar?
Justyna: At the beginning of my career, it was a challenge to be seen as more than a blonde. I had to work harder to show that it wasn’t my looks — but my brains — that made me a competent and productive team member and lead.
It doesn’t have to be like that. And I’m happy to see that at some professional communities I don’t feel judged by my gender but rather by my content and my ability to do what I contribute. It typically takes less than five minute technical conversation to get to that level of respect.
“If girls have role models, like technical women, it helps them answer the question: who do I want to be in the future? What is the model that I want to follow.” — Justyna Zander
And what are your responsibilities as the Ecosystem Lead at Intel?
Justyna: I am responsible for building the architecture of a unified software ecosystem required for autonomous driving. The software, chips, and integrated computing platforms I’m working on will enable the AI algorithms required for a safe and secure autonomous driving platform.
What intrigues you most about autonomous driving?
Justyna: I find the complexity intriguing. Autonomous driving requires embedded system domains, electrical engineering, computer science, simulation, even mechanical engineering and so much more. It’s interesting to work towards the answers to questions such as: How is it going to work on multiple technical levels? What’s the compute power required? How do we program this so that the power efficiency, technology and everything else allows for the required ecosystem to work? The complexity of the learning curve is fascinating.
What is the problem you are trying to solve and for whom?
Justyna: I’m working towards solving a problem for the auto industry. You see, the auto industry is heading toward a driverless cars model but there are variety of tech gap problems to scale things up. Today, we are still figuring out the powerful computing platform designs that allow for optimal, scalable, power efficient, safe and secure autonomous cars and autonomous car ecosystems. I’m working to getting us there through technology.
I’m working through the kind of software to make full autonomous driving safe and secure; figuring out the design needed for a model of car and an ecosystem that can be commercialized and go global.
Can you describe the tech ecosystem that is required for self-driving cars?
Justyna: An autonomous driving ecosystem requires these 4 trends and more…
- Trend 1 — Autopilot aka Advanced Driving Assistance System: Multiple variations of it have been around for at least a decade.
- Trend 2 — Artificial Intelligence and Deep Learning: think of Google’s DeepMind but for embedded systems.
- Trend 3 — High Definition mapping: we now have sensors and 3D mapping technology. You can imagine it as a real time video-stream from a car to compare with the map masterfile. If there is a mismatch, the car’s computer/server on the wheels reevaluates its driving.
- Trend 4 — Cheaper sensors and edge-computing chips: this makes it easier for auto manufacturers to turn their cars into sensing computers in a cost-effective manner.
Up ahead, we still have to figure out power efficiency, functional safety, high definition mapping of the planet, and verification and validation of data, deep learning algorithms,and actions performed. This is where the understanding of physics, simulation, and deep learning integration and interpreting these cross-sections properly.
What role does AI and cognitive computing play in self-driving car technologies?
Justyna: AI and cognitive technologies play in multiple dimensions in the future of mobility and autonomous driving field.
There is the literal deep learning aspect. This is really important, as this is the technology in a car that is responsible for taking images and videos streams, analyzing what’s happening frame by frame, and then cueing the car to take action. In short, AI allows for the car to learn and to become more accurate with time.
But beyond that, AI and cognitive technologies play a huge role in the future model of autonomous driving where the car does more than drive from point A to point B. In the future, with the help of cognitive technologies a car will provide a new mobility experience. The user of a car will communicate with the car and visa versa. There will be a fetching mechanism. The car will know your schedule, be your assistant, figure out optimal ways of navigating your day, provide you with tips, answers to your questions, and ultimately provide you with a personalized journey.
What role does AI, the cloud, and Internet of Things (IoT) play in self-driving car technologies?
Justyna: AI, the Cloud and IoT are essential for the future of autonomous driving era. For example, a fleet of driverless cars is really a fleet of IoT devices, being instructed by AI that could run in the cloud in a smart city.
It is only with the converging technologies that we can optimize traffic, optimize the process of battery charging, and get optimized services for you. It is with the power of AI, the cloud, and IoT that your future driverless car experience will evolve from being a vehicle that takes you from point A to point B to a vehicle that takes you on a journey that fits your personalized schedule.
But before we get there, we need to first figure out the autonomous vehicle, the ecosystem, and then focus on the cloud.
How far are we from figuring it out?
Justyna: Depending on who you ask, it can be 2 to 25 years before we figure it out.
We’ve achieved advanced driving assistance system level. That’s just the beginning. Only a few provide true self driving capabilities, Google car being the example, but not on a commercial scale.
“There is a distinct difference between true self-driving cars and advanced driving assistance systems or autopilots.”- Justyna Zander
Getting to commercialized and full autonomous driving will be a longer process. There is so much to do. Some of it includes: achieving acceptable levels of process validation, making sure we reach ISO 26262 certification, and creating business models that allow for auto manufacturers to adopt autonomous driving technology cost-effectively.
The adoption of the autonomous driving model will be yet another challenge. It’s going to take time to change human behaviors, replace cars, and modify the mobility services.
What are your predictions for autonomous driving?
Justyna: Autonomous driving will be gradually integrated into the world. The technology required for it will get cheaper. There will be immense progress done in building in the intrinsic functional safety and security.
What are your favorite driver-less car use cases?
Justyna: It’s exciting to think that one day we will be in a world where there is no need for parking or traffic congestion.
The transportation and storage of goods will be optimized.
We will capitalize on the autonomous driving services by building new business models. Can you imagine, your car working for you? One day, your car will drive you to your destination and then get to work driving people around. Your car may make you financial revenue and help you get around in a more mindful manner.
How can we learn more about you?
Justyna: You can visit my personal site here.
Opinions expressed in this article represent Justyna Zander’s views and are not the views of her employer.
Originally published at www.huffingtonpost.com on September 21, 2016.