Anika Ramachandran standing in front of the doors of Etcheverry Hall.

With her recent experience working as a Software Engineer at Nuro, Anika Ramachandran (BS IEOR ‘22) has spent much time immersed in the world of autonomous vehicles. We recently caught up with Anika to hear her take on what the future holds for self-driving cars.

1. Could you share any notable advancements or innovations in self-driving technology in the past few years?

One significant advancement is the improvement in sensors. Both the accuracy and range of sensors, such as cameras and lidar have increased. This enhanced sensor technology allows self-driving vehicles to perceive the world better and make more informed decisions. Another noteworthy development is the pivot towards AI/ML (Artificial Intelligence/Machine Learning) as the major approach in self-driving technology. In the past, planning was often scenario- based, but with the advancements in deep learning, teams are now adopting AI/ML methods to power the software behind self-driving vehicles. There has also been a rise in the number of real- world pilots on roads. This means that self-driving technology is being tested in practical scenarios, giving us insights into how it could work commercially. This also helps gauge the public’s attitude toward autonomous vehicles.

2. In your opinion, what are some of the key benefits that self-driving cars can bring to society, and how do you envision them transforming various industries?

I believe that self-driving cars have the potential to bring about significant benefits to society and revolutionize various industries. First and foremost, these vehicles can be designed to prioritize safety, drastically reducing accidents by eliminating human errors that often lead to traffic incidents. The proper implementation of self-driving fleets can significantly enhance mobility in cities, offering convenient transportation options for individuals who cannot drive themselves. These services can be provided more safely and at a lower cost than traditional ride-hailing services because human drivers are no longer needed. By optimizing vehicle utilization and reducing the demand for parking spaces, cities can also benefit from better traffic management and less congestion. The potential for increased efficiency and optimized traffic flow is also significant.

Self-driving cars can communicate with each other, leading to even greater efficiency and optimized traffic flow. Another noteworthy advantage is that self-driving cars can save valuable time for passengers, allowing them to be more productive during their journeys. This newfound productivity could have a positive impact on various aspects of life, including work and leisure. Self-driving technology can also contribute to environmental sustainability. By employing optimized driving patterns, these vehicles can minimize fuel consumption and carbon emissions, making them more eco-friendly compared to traditional cars.

However, it is essential to address the valid concerns surrounding self-driving technology, such as potential job loss in the driving industry and occasional accidents that make headlines. These concerns should be taken into account during the development of the technology. The positive impact of self-driving cars on society, safety, mobility, efficiency, and the environment cannot be ignored; however, as we continue to advance in this field, it is crucial to strike a balance between progress and responsibility, ensuring that this technology can bring about the transformation it promises while addressing its challenges in a mindful manner.

3. As self-driving technology continues to evolve, what do you believe are some crucial areas that require further development or research to ensure the safe and widespread adoption of autonomous vehicles?

Most companies already recognize this, but safety must be at the forefront during the development of this technology. Models and sensor stacks need to be refined to instill high confidence in the ability of self-driving vehicles to perceive the world accurately. While humans also make mistakes, the ultimate goal is for self-driving to eliminate “human error” entirely.

Additionally, there is a need to catch up on the regulatory front. Laws governing the safe operation and development of self-driving vehicles are being created in parallel with the advancement of technology, which means it’s not always clear what regulations are needed. It is essential to be predictive and ensure potential cybersecurity and safety constraints are in place before widespread adoption.

Another critical aspect to develop for widespread adoption is the capability of vehicles to navigate previously unseen areas. The world is constantly changing, and self-driving systems must be able to adapt to new environments. Planning systems can’t perfectly predict far into the future. Thus, testing should go beyond specific regions to understand how self-driving stacks perform under different driving conditions in diverse areas.

4. In your experience working with autonomous vehicles and AI, what are some of the notable benefits and limitations of AI technology in enhancing the safety and efficiency of self-driving cars?

One benefit AI has enabled is the development of better perception models so self-driving cars can better understand their surroundings. AI also allows for greater adaptability- self-driving cars have continuous learning capabilities, allowing them to improve their decision-making skills by learning from new scenarios encountered on the road. Finally, AI enables self-driving cars to process large amounts of sensor data from cameras and other sources to facilitate real-time decision-making and navigation. 

Limitations include potential challenges if self-driving cars encounter unforeseen and rare road conditions, as they heavily rely on the data they have been trained on. The lack of individual scenario control in AI systems can also introduce ethical ambiguity, as decisions made by self-driving cars in certain situations may not have clear-cut solutions, raising ethical concerns.

5. With the growing concern for environmental sustainability, how do you envision self-driving cars contributing to a greener future?

I think self-driving technology has the potential to make significant contributions to a greener future both as individual vehicles and if we consider their broader impact as fleets. Besides most self-driving cars being electric or using alternative fuels, they can leverage real-time data to identify efficient routes and optimize acceleration/deceleration thereby avoiding wasted energy. It can also reduce traffic bottlenecks and provide an opportunity to eliminate unnecessary trips, which can contribute to reduced overall carbon emissions. Finally, by enabling shared autonomous vehicle services, self-driving can also help reduce the number of privately owned vehicles on the road, leading to reduced traffic congestion and lower carbon footprints associated with vehicle production.

This article originally appeared in Berkeley IEOR Magazine.