paint-brush
Isht Dwivedi Is Working Toward Developing Safer Next Generation Carsby@craiglebrau
312 reads
312 reads

Isht Dwivedi Is Working Toward Developing Safer Next Generation Cars

by Craig LebrauNovember 2nd, 2022
Read on Terminal Reader
Read this story w/o Javascript
tldt arrow

Too Long; Didn't Read

Honda has set a far-reaching goal of cutting traffic collision fatalities to half by 2030 and eliminating them completely by 2050. Isht Dwivedi is one of these engineers at Honda who dedicates each day to advancing computer science techniques. He is the inventor of multiple patented systems and techniques and has even had his work on road condition monitoring featured as a video on the official Honda YouTube Channel. The challenges that Isht will face in his career are complex and difficult, but with perseverance as one of his most prominent qualities, he's up for it all.

People Mentioned

Mention Thumbnail

Companies Mentioned

Mention Thumbnail
Mention Thumbnail
featured image - Isht Dwivedi Is Working Toward Developing Safer Next Generation Cars
Craig Lebrau HackerNoon profile picture

According to the National Highway Traffic Safety Administration’s (NHTSA) projections, an estimated 42,915 people died in motor vehicle traffic crashes in 2021, a stark 10.5% increase from the 38,824 fatalities the year before. In order to mitigate such tragic loss of human life, Honda - one of the world’s largest automakers has set a far-reaching goal of cutting traffic collision fatalities to half by 2030 and eliminating them completely by 2050. Setting such ambitious targets would not be possible without persistent efforts by research engineers designing the technology and tools to bring them to reality. Isht Dwivedi is one of these at Honda- he dedicates each day to advancing computer science techniques, such as researching new computer vision methods to figure out how to make cars more autonomous and roads safer.

As a child, Isht was fascinated by the possibility of automating complex tasks such as driving a vehicle. He knew that technology could be used to make processes easier, so he began to take an interest in engineering. When it was time to pursue a university education, Isht studied courses related to Artificial Intelligence (AI) at the Indian Institute of Technology in Guwahati, a top-ranked technology university in India. He explored a plethora of AI & automation-related topics such as machine learning, computer vision, pattern recognition, and more. For his thesis, Isht developed an algorithm to analyze handwritten documents to determine the identity of the writer. He achieved state-of-the-art prediction accuracy with his work, a feat highly valuable in forensics.

Isht went on to work at the Indian Institute of Science, but once he was there, he felt he needed to take more time learning the complex skills he wanted to apply in his career. So, he went back to school at Columbia University to pursue a master’s degree in computer science. Equipped with a refined set of technical skills, Isht was hired by the Honda Research Institute to tackle projects related to Advanced Driver Assistance Systems (ADAS) and Autonomous Driving Systems (ADS) within the United States.

At Honda, Isht is leading a project focusing on road scene understanding. Using road scene video classification and captioning technology from cameras mounted on a moving car, the developed technology can determine road surface conditions, road type, etc., and provide detailed information about other variables in the scene such as construction zones or risky agents on the road. This work is being used in Honda’s development of next-generation interfaces to provide personalized navigation support to drivers.  This system will assist the driver in safety-critical and convenience applications by serving as an agent that communicates information. 

In another project, Isht is helping Honda steer away from dependence on High-Definition (HD) maps that many current autonomous driving systems depend on. HD maps contain the precise location of everything in a scene such as stop signs, traffic lights, lanes and lane lines, but these maps have limitations - they have to be developed individually and updated frequently, making them difficult to scale. With advanced technology using computer vision, these maps can be replaced by Bird’s Eye View (BEV) segmentation that relies only on video data from camera sensors on the car. BEV segmentation aims to build an accurate map of the car’s surroundings. When completed, this will be the new approach to autonomous driving because it allows cars to adapt to dynamic variables and is much more scalable than the current mechanisms in place. This technology is vital when it comes to the development of both ADAS and ADS at Honda. 

Isht’s research is not only integral to the success of Honda, but it’s also pivotal to the advancement of the automotive industry as a whole. He is the inventor of multiple patented systems and techniques and has even had his work on road condition monitoring featured as a video on the official Honda YouTube Channel. The challenges that Isht will face in his career are complex and difficult, but with perseverance as one of his most prominent qualities, he’s up for it all. Since his childhood, Isht has seen the potential applications of technology in humanity’s everyday life, and now, he gets to make that vision become a reality while making driving safer, too.