Photo for representation
Because driving behaviour varies significantly across regions, even when traffic rules are similar, senior executives of Thyssenkrupp Automotive Technology said, drawing parallels with the approach followed in Europe, pune: Advanced driver assistance systems (ADAS) need to be tested and trained separately for different countries.
Thyssenkrupp currently has around 120 engineers at its Pune engineering competency centre and plans to increase the workforce to more than 300 in less than a year. Because most of the data that was used to develop ADAS was based on European traffic conditions, automotive Research Association of India (ARAI) recently developed a test track to create data for ADAS suited to the Indian road conditions. Because you get real-life or real-time data collection, then you can use those data to increase the speed of implementing those ADAS systems in the country, “I think with using the data in India,. Because the latter offered greater flexibility, erbe said conventional rule-based algorithms were increasingly being outperformed by AI-based systems.
You Can Also Check: Gold Rate in Pune | Silver Rate in Pune | Bank Holidays in Pune | Public Holidays in Pune | Petrol Price in Pune | Diesel Price in Pune | CNG Price in Pune | LPG Price in Pune Stay updated with the latest Pune news.
Viktor Molnar, COO, Thyssenkrupp Automotive Technology, said automotive technology developers already train ADAS applications to account for country-specific driving patterns in Europe, the US and Asia. But maybe the challenge is the amount of data you will need to train the algorithms and to make them competitive is larger,” said Bjorn Erbe, senior VP of R&D Bilstein Motion Solutions. India’s complex traffic presented an opportunity not just for govt research institutions but also for startups to develop solutions that could be commercialised and made more region-specific, Erbe said. Molnar said differences in driving behaviour were evident even between European countries, making local adaptation an important part of automated driving technology development. And exactly this decision-making is something that you have to basically recreate in your automated driving system,” he said. Otherwise, you are ending up in an endless traffic jam,” Molnar said. Molnar said collecting more data on Indian traffic would help technology developers accelerate the introduction of ADAS applications suited to the country. And I think that’s a very valuable set of data,” Molnar said. Erbe agreed with Molnar’s assessment and said the scale and complexity of Indian traffic would make the availability of adequate training data particularly important. The executives said artificial intelligence (AI) could further transform the way ADAS applications are developed, particularly in complex traffic environments. “Normally, you have rules and then you train algorithms according to certain datasets, but the rules are static. When you bring AI, you replace these rules with artificial intelligence applications, which are much more flexible,” he said. He said the greater the complexity of traffic, the more relevant these systems would become, adding that India could give a further boost to this development. “With AI, you can combine different data sources, whether it’s coming from a car or coming from the infrastructure,” he said, adding that environmental data could help train algorithms more efficiently and quickly. Molnar said automated vehicles would have to understand and accommodate human driving behaviour, while human motorists would also need to accept automated vehicles.
India’s particularly complex traffic conditions would require larger volumes of real-world data to make algorithms competitive, but could also create an opportunity for local startups to develop and commercialise region-specific solutions, he added. “Any ADAS application has to be trained for certain region specifics. You also see it in Asia and the US. It is a business unit of Thyssenkrupp. “Even in Europe, country by country, it is very different how drivers react and make decisions. Molnar explained that an automaker needed to capture different real-world driving situations to understand how motorists respond and train automated systems to replicate those decisions. “For example, you guys probably go into situations where you would just apply the brake. And then this is exactly what you have to consider. If you automatise, then you need to train the algorithm to make exactly the decision that people are doing here. differences in how motorists behave make region-specific datasets essential While traffic regulations may have similarities across markets. He added that real-time data collection could provide valuable inputs for developing and implementing ADAS technologies in India. The experience in Europe had also been evolutionary, with developers progressively adapting systems to individual markets. He also saw an opportunity for Indian startups to build specialised solutions around local traffic conditions rather than limiting the effort to institutional research. The executives also pointed to the challenge of ensuring that human-driven and automated vehicles coexist on roads during the transition towards higher levels of vehicle automation. Download the TOI App.