Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Autonomous driving through extremely-tight-dynamic environments with complex, stochastic, and adversarial traffic-dynamics, or simply through an absolute chaos, on sub-urban unstructured roads in India. This kind of traffic and environment has never been attempted in the history of #autonomousdriving. There were no traffic-rules to abide by on this road, other...

125,317 Aufrufe • vor 2 Jahren •via X (Twitter)

10 Kommentare

Profilbild von Qui Gon Jinn ✨⚡
Qui Gon Jinn ✨⚡vor 2 Jahren

Appreciate the work but it doesn't seem like those other people in traffic had knowledge of what you were doing and I doubt they at least I wouldn't appreciate being a unwitting participant in your test drive. What if your system misjudged something and throttled hard on the accelerator?

Profilbild von Sanjeev Sharma
Sanjeev Sharmavor 2 Jahren

It is more of a risk for us than to anyone else. It is career at stake. We wouldn't do anything for a cheap fame. Unless the tech is rigorously tested, we wouldn't do anything. Yesterday media came to do live demo off-roads. We realized there was a construction going on with pile of mud dumped from dumper trucks on the side of the road. I gave them an explicit warning, that I know our vehicle will avoid it but this is untested. They said their safety inside our vehicle is their risk, and only then we did the demo off-roads, and did two tests, where our vehicle did avoid the pile of mud on the side, before doing a live demo with them. The vehicle was never shown before pile of muds, and our explicit obstacles detection software was also never trained for that. So whatever we do, safety is of paramount importance. Everything else can wait in life. This motion planning and decision making framework has gone rigorous testing. Our electromech system is designed as such that a safety driver can easily take control. In Bolero it is controlled by a motor which has to actuate and thus there is no electrical current based system that cannot be controlled or monitored. This is the reason why we are not running Thar in dense roads, because it has directly input to the ECU.

Profilbild von Akshay Charegaonkar 🇮🇳
Akshay Charegaonkar 🇮🇳vor 2 Jahren

This is miraculous...your autonomous vehicle is already better than 90% of Indian drivers 🤣

Profilbild von Anirban Roy Das
Anirban Roy Dasvor 2 Jahren

When you think of driverless vehicles, one thing that never comes to our immediate attention are the roads. We are so used to seeing autonomous vehicles run in controlled environments or great quality roads. Our imagination of self driving cars are always picturing a developed nation with HQ roads, no potholes, no uneven surfaces, at least not at every 10 meters. Seeing autonomous vehicle doing path planning on Indian roads, and in this video specifically, it is as basic as it can get in terms of majority of Indian roads, is highly appreciable. Even though there are so many more variables are at play and many more things to do, this gave me such a positive kick.

Profilbild von Himanshu Gaurav Singh
Himanshu Gaurav Singhvor 2 Jahren

Great demo! At 3:11 in the video, there is a person visible on the driving seat. Does that person manually operate the car at any point of time? If yes, then could you please elaborate which of the manoeuvres were human-assisted? Thanks in advance! Really great work!

Profilbild von k
kvor 2 Jahren

It is doing lot of last minute obstacle avoidance. More like swaying away. Could this be a feature or you working on to make it anticipate incoming cars much faster when they are at a safer distance?

Profilbild von 2cents
2centsvor 2 Jahren

Genius

Profilbild von Shivam
Shivamvor 2 Jahren

I believe the car's vision is able to find the traversability region and utilise it... but its still inefficient when we got a traffic scenerios and a sudden appearance of say a bike(occlusion).. that's why we are still long away from having a driverless car

Profilbild von VSidd
VSiddvor 2 Jahren

Cool stuff. Really appreciate the people who did this work 🙏🏻

Profilbild von We r🤘🤘
We r🤘🤘vor 2 Jahren

Awesome never seen this before and simply love 😍the Flag on the vehicle🚩🚩🚩🚩

Ähnliche Videos

Autonomous driving through tight, dynamic, stochastic, and adversarial traffic-dynamics on sub-urban roads in India, as well as through partially unstructured environments. This demos showcases the robustness of our motion planning and decision making algorithmic frameworks in enabling #autonomousdriving through seamlessly through such traffic and environmental scenarios. The vehicle starts from a generic open environment at the temple, where there are no traffic-rules to abide by. It then exits the region and assumes a generic autonomous navigation behaviour, negotiating complex traffic scenes. At various points it can be seen that the other vehicles (bikes, autos, bicyclists, and cars) didn't abide by any traffic-rule and moved in crisscross fashion, presenting adversarial scenarios, challenging our autonomous vehicle at Swaayatt Robots to take care of the collision avoidance. This classical motion planning and decision making algorithmic framework is being further scaled up with deep #reinforcementlearning, which will practically solve the sub-urban traffic-dynamics and environment negotiation for #autonomousvehicles in India and throughout the world as well. This demo was done at the Kankali Kali Mata mandir in the city of Bhopal. This demo was a culmination of our prior works and demos: off-roads, on-roads, bidirectional traffic negotiation in single lane roads, and toll-plaza navigation. We have taken up the arduous task of solving the Level-4 autonomous driving by the end of 2024, globally. #machinelearning #deeplearning

Sanjeev Sharma

186,291 Aufrufe • vor 2 Jahren

Autonomous driving through very dense dynamic traffic, with extremely tight-complex-stochastic traffic-dynamics on sub-urban roads, connecting to an open ground, with absolutely zero traffic-rules. This is the most heavily cluttered environment where we have tested our #autonomousdriving technology, presenting many of the adversarial negotiation scenarios as well, throughout the autonomous navigation task. This demo was done at the Mata Baglamukhi Madir campus in the city of Nalkheda, in MP, India, and was done in the presence of heavy police forces deployed that day on the ground, as can be seen in our demo. Our autonomous vehicle starts from the temple with a generic open environment, with zero traffic rules, with very narrow corridors created out of barricades for vehicles movement by the security forces. In the corridor no two vehicles can pass through at the same time, and our vehicle was tasked with driving through this corridor, while negotiating its way from any traffic, two-wheelers, or pedestrians it faces, with dense presence of bikes and cars on either side, presenting a very challenging environment for #autonomousvehicles. The vehicle exits the open area, and then assumes generic dual lane navigation, avoiding both static and dynamic obstacles, before encountering a police check-post, where the vehicle is supposed to wait if the barricade is closed, and proceed if open. Upon exiting the checkpost, the vehicle negotiates a traffic-intersection with stochastic and adversarial driving behaviour of other vehicles on the road. Our vehicle continuously faced heavily cluttered traffic scene, where entities on the road can execute a random driving pattern, making the decision making task very challenging. We did the demo over a period of two days, successfully executing multiple (30+) trials in this setting. This demo was again a culmination of our prior works and demos: Kankali Kali Mata demo, on-roads, bidirectional negotiation capability on single lane roads, and open environment Level-5 negotiation capability as showcased in our Toll-Plaza demo. We again scaled up classical decision making and motion planning algorithmic framework, to adapt to such a level of density of obstacles on the road. This framework is further being scaled up with #reinforcementlearning and unsupervised #deeplearning at Swaayatt Robots. We will again do a demo in the month of June here, showcasing autonomously acquired skills to pave the way for Level-5 autonomous driving, and to solve the Level-4 autonomy problem by the end of 2024. #MachineLearning

Sanjeev Sharma

332,976 Aufrufe • vor 2 Jahren

Autonomous vehicle learning to dodge traffic, performing stochastic adversarial negotiation. On 27th August we had representatives from the Suzuki Motor Corporation's autonomous department, Genki Maeda (Department Manager, AD Platform Development), Karachi Nobunari (Department General Manager, Advance Technology Development Department) and Ronit Kumar (Suzuki Innovation Center) visit us to test our #autonomousdriving technology. This was a high-stakes demo, where we asked our engineering team (including the founder, Sanjeev Sharma) to ride two wheeler and to cut the path of our autonomous vehicle at random / at will, in a live demo, creating an adversarial scenario, where it is the sole responsibility of our autonomous vehicle to dodge obstacles and prevent accidents. Over the years, we have been building autonomous driving technology to enable negotiation of adversarial-complex-stochastic traffic dynamics. This demo is only a short trailer of what is being developed and what is going to come next. We made the vehicle first negotiate randomly placed static vehicles, bikes and cones on the road. Then in the next section of the road, our engineering team started cutting the path of autonomous vehicle at random, and let it take care of obstacles avoidance and negotiation, balancing aggressiveness and passivity. These algorithmic frameworks are being scaled up to achieve Level-4 and Level-5 autonomous driving, in the most complex of traffic situations imaginable in the world for #autonomousvehicles, i.e., Indian traffic on Indian roads, to conquer this space globally. The speed of the vehicle was kept low in this demo, keeping in mind the safety of the guests. The core motion planning and decision making algorithm in the demo utilized one #reinforcementlearning agent. There is going to be another demo on similar lines next week, on an extended stretch of a road. #deeplearning #India

Swaayatt Robots

22,375 Aufrufe • vor 1 Jahr

Presenting autonomous driving in complex, stochastic and adversarial traffic-dynamics, on the roads in #India. Over the course of last year, we enabled #autonomousdriving in conditions and in situations no one in the autonomous driving industry considered was possible, and in a country like India, where the contemporary belief was that #autonomousvehicles are an impossibility. With the closure of 2023, we present autonomous driving at a very large scale, in the city of Bhopal in India. In this demo our autonomous vehicle at Swaayatt Robots can be seen negotiating the surrounding traffic with ease, where it had to deal with their stochastic and adversarial driving patterns, like suddenly switching lanes and appearing all of a sudden in our vehicle's current driving lane, without adherence to the traffic rules. This demo was a culmination of the cutting-edge research we have been doing, and the technologies we have been developing, over the years, which we showcased throughout in our demos in 2023 -- campus autonomous driving (February), off-roads autonomous driving (April and September), tight-stochastic traffic negotiation (August and September), bidirectional traffic negotiation on a single lane road (October), large-scale city level demo (November), and Toll-Plaza negotiation (December). In 2024, we will scale our technology commercially, and bring it to North America and Europe, to topple this trillion dollar industry, and will also scale our technology throughout India. Wishing everyone a very Happy New Year! #deeplearning #reinforcementlearning #MachineLearning Elon Musk PMO India Narendra Modi Nitin Gadkari DARPA DRDO

Sanjeev Sharma

95,684 Aufrufe • vor 2 Jahren

In this demo we extend our prior work on obstacles avoidance at aggressive speeds, showcasing our Thar based autonomous vehicle navigating at near drift speeds, progressing towards our endeavour of Level-5 autonomy. Our autonomous vehicle at Swaayatt Robots was tasked with avoidance of traffic cones on the road, placed in a zig-zag fashion, at aggressive speeds. The location of the marked cones was not known to the planner beforehand. The #autonomousdriving task, i.e., motion planning (time parametrized trajectory computation) and decision making, was made even more challenging by restricting the AI agents to not act on obstacles unless they are within 24m radius. Level-5 #autonomousvehicles should be able to react quickly to overtake, or to avoid, any sudden unforeseeable obstacle or pedestrian on the road to avoid fatalities -- a capability demonstrated by our novel motion planning and decision making algorithmic framework over here. Our previous demo showcased our Bolero based platform consistently keeping speeds beyond 45 KMPH for most part, slowing down to only 39 KMPH at one point. Given Thar has lesser body roll, our framework successfully kept speeds well above 47 KMPH (even at the points of avoidance of obstacles), with speeds reaching as high as 55 KMPH. A typical human driver would feel uncomfortable at speeds beyond 40 KMPH in such as scenario. The entire algorithmic framework with 5 classical (one #reinforcementlearning-) agents , runs at 800+ Hz on a regular i7 processor, single thread. This algorithmic framework is being further scaled up with end-to-end deep reinforcement learning, and will be showcased in the month of March. #deeplearning #machinelearning #motionplanning

Sanjeev Sharma

12,841 Aufrufe • vor 1 Jahr

Danger on Duty: Barrier Dragged on Bike in Busy Traffic How are road barriers being moved on Bengaluru roads? In a recent video, it can be seen that traffic personnel are shifting a road barrier in a very unusual and unsafe manner. One person is riding a bike slowly, while another person is holding the barrier by hand and moving along with it, matching the speed of the bike. He is not exactly transporting the barrier on the vehicle, but rather dragging or guiding it while walking or running alongside, using the bike’s movement to keep it in motion.Even though the intention may be to reposition the barrier quickly, this method raises serious safety concerns. Handling a heavy road barrier in such a way requires precise balance and coordination, and any slight misjudgment can cause the barrier to slip, fall, or collide with nearby vehicles or pedestrians. On busy Bengaluru roads, where traffic conditions are already unpredictable, such practices significantly increase the risk of accidents.Moreover, this kind of improvised handling does not reflect proper safety standards. Road barriers are meant to be moved using appropriate equipment or vehicles designed for such tasks. Relying on a two-wheeler and manual handling not only puts the individuals involved at risk but also endangers everyone else on the road.Incidents like this highlight the need for stricter adherence to safety protocols and better operational practices. Public safety should always be the top priority, and such risky methods must be avoided in favor of safer and more responsible alternatives. #bangalore #bengaluru #barricade #bengalurupolice ಬೆಂಗಳೂರು ನಗರ ಪೊಲೀಸ್‌ BengaluruCityPolice ಬೆಂಗಳೂರು ಸಂಚಾರ ಪೊಲೀಸ್ BengaluruTrafficPolice CP Bengaluru ಪೊಲೀಸ್ ಆಯುಕ್ತ ಬೆಂಗಳೂರು Joint CP, Traffic, Bengaluru alok kumar DGP KARNATAKA ಕರ್ನಾಟಕ ರಾಜ್ಯ ಪೊಲೀಸ್ Karnataka State Police S. Lalitha ChristinMathewPhilip

Karnataka Portfolio

13,418 Aufrufe • vor 4 Monaten