A Drive Without A Driver

Automatic Summary

Discover The Intriguing World of Self-Driving Car Technology

Today, I wish to introduce you to the fascinating realm of self-driving cars and autonomous vehicles. As Dave Waters aptly put it, "Once you trust a self-driving car with your life, you pretty much will trust artificial intelligence with anything." Self-driving cars aim to remove the burden of driving from our daily lives and create safer, more efficient roads.

Understanding Advanced Driver Assistance Systems (ADAS)

Advanced Driver Assistance Systems (ADAS) are at the core of self-driving technology. These systems aim to increase safety and comfort by automating specific driving tasks. They can be broadly divided into Active Safety features and Passive Safety features.

  • Active Safety features aim to prevent crashes or at least, reduce the severity of crashes.
  • Passive Safety features, on the other hand, protect the vehicle and its inhabitants once a crash has already occurred. For instance, airbags and seat belts are key passive safety features.

The Role of Sensors in Self-Driving Technology

Sensors are integral to making self-driving cars a reality. Various sensors such as long-range RADAR, LIDAR, cameras, short to medium range RADAR, and ultrasound sensors capture essential information. This data then powers ADAS features like intelligent speed assist, which adapts the vehicle's speed according to recognized road signs.

Between Alerts and Actuation

Based on the sensory input and software, a self-driving car can either alert the driver about potential issues or take control to avert accidents. This categorization leads to two broad types of scenarios: Alerts, which signal the user to a potential threat, and Actuation, where the vehicle takes proactive control to ensure safety.

Levels of Automation in Self-Driving Cars

Automation in self-driving cars varies significantly, right from no automation to full automation.

  1. No Automation: The human driver takes full control of the vehicle.
  2. Driver Assistance: Self-driving technology assists the human driver without taking full control.
  3. Partial Automation: The vehicle can steer and control speed by itself, but under human supervision.
  4. Conditional Automation: The vehicle takes complete control in specific conditions.
  5. High Automation: The car takes full control regardless of human intervention.
  6. Full Automation: The vehicle performs all driving tasks under every condition.

Future refinements in AI and sensor technology will continue to push the boundaries, moving us closer to true, full automation.

Final Thoughts

While the concept of self-driving car technology may seem futuristic, it's closer than we think. However, achieving high or full automation requires substantial infrastructure changes, something not all countries are equipped for at the moment.

Increasing the degree of automation in cars not only promises to enhance human safety and convenience but also represents an exciting advancement in artificial intelligence and machine learning capabilities. Explore this fascinating topic further, and join in this technological revolution!

Thank you for joining this journey through the world of self-driving car technology and its impact on our lives.


Video Transcription

So right now I'm having the session to talk to you something about the self driving car and like where we are right now in this autonomous vehicles and a little bit of things like that.So summary, like uh Dave Waters once said something like uh once you trust a self driving car with your life, you pretty much will trust artificial intelligence with anything. Like for most of the people driving is a difficult task like in the traffic and like uh and everything like driving is a part of their life and it is a bit difficult task and this autonomous driving vehicles and things like that is actually helping people to make their lives better.

So uh a little bit of introduction about myself, I'm at the level stretch. I'm working as a senior engineer in transportation B of Data Galaxy and uh I'm specialized in electric vehicles and technology. So like there may be times uh times when you stand and look around and feel like if everything is automated, if everything is autonomous, things will be like a bit easier for you. Like you don't have to put an effort in anything you don't like uh somebody maybe uh liking to drive, but there may be people that don't actually want to drive. So for people like that, this autonomous vehicle and uh this A I systems is actually a very helping feature that is actually growing uh in an exponential way in like in past few few years. So let me start with this big concept like uh what does this, how this or this driver assistance systems came to their life? So this autonomous vehicle or basically we call it as s technology. That is nothing but this advanced driver assistance system. Actually, this assistance systems came because uh we need to increase the safety. Like, you know, you may be having this uh you may be uh victimizing something like accidents or anything you want to protect your family or you want to protect yourself.

So you may have to add something like which is more safe. So if you're driving with your kids or something like that, we want, we don't want to hurt anyone. So in situations like that, we always try to improve the hm place we're living the vehicle we use and everything like that. So a a is a technology that came for us like uh basically, it started with a concept called active safety and passive safety. So what active safety actually is, is like it will prevent like prevent something from happening, prevent crashes or reduce the severity of the crashes or things like that it is the part of this active safety and for passive safety, what it is like, it will protect the vehicle or it will protect the occupants of the vehicle.

Once the crash has already occurred, something like airbag and seat belt and things like that, this airbag actually, you know, like it to make it implanted, the crash has to get happen. So things like that come in passing safety features and in active safety, we are having a much more diverse set of features. I will introduce you some like uh some terms and systems and that we currently use uh in day to day life. So uh maybe you'll be hearing some terms like this as AC C that is to control and this forward collision warning, high beam safety system, lane departure, warning, traffic signal, detonation, traction control, urban road speed management in intelligence speed, like these are only a group of like systems that is actually giving us uh a way to achieve this active safety.

So you can see a picture here. Uh like this is a of a card uh which is implemented with some speeches. So we'll be having different set of sensors like you can see here long range shredder, late R camera, short medium range shader, ultrasound sensor, something like that will be occupied in your vehicle. And it will be giving us the details or it will be capturing a lot of details uh that is required to make this all these systems work. Like maybe uh if you need to get an example, something like that. Uh I work with a particular thing called intelligent speed assist. It is nothing. But if you see some roads, road sign limit or something, we have to adapt the speed accordingly. Hm. So while you're traveling on the road, you see a speed limit, you will be able to process that, but the vehicle should also be able to process that if you're automating this process, right? So for that, actually, we will be having some particular sensor, you will be basically using cameras and red eyes or things like that and we will be uh taking the picture and uh like what exactly the road sign is if it is a stop sign or if it is a speed limit or things like that, it will get that information, it will collect the information, it will give the system what exactly the speed limit is or what exactly the sign means.

And the information is passed to the main system and uh our maximum speed or the speed limit value will be set accordingly. So something like that is happening in intelligent, the so in every features like in every each and every system I mentioned here, there are a lot of happening inside like, so basically, uh we need the sensor input, which is very important. I will sh uh show you like uh how an A I system looks like. So it is basically this, we'll be having consensus and uh we'll be having this particular at SECU which is implemented with the software that we are working on. Uh like maybe it can be AC C it can be intelligently resistor, it can be anything. So this will be implemented in this particular hardware and we'll be having the sensor input coming to this and it will process the information and it will give the alerts or actuation like uh two kinds of situation can happen. Something like uh there are things like give you some warnings like you are exceeding speed limit or something like that is called alerts. Like it will just tell it the user.

But there are some actuations also happening like the vehicle itself will control the process will uh the vehicle itself will adjust the speed, the vehicle itself will uh like adjust the lane which is traveling or things like that can also will happen. Like if you have to like uh break suddenly because there is a passing uh pedestrian, passing things like that, even this vehicle can take control by itself. So uh that is basically about this alerts and actuation that is coming from the esecu. So basically you'll be having some sensor and you'll be having this hard hardware platform and interfaces and then we'll be having, which is implemented with the software algorithm like whatever it can be the AC C or anything. And then we'll be like colors and then there can be, this actuation actuation is assisting our full control of the vehicle manually. Like it can, it's self control. Like whether we should take a turn, whether we should uh turn on a lamp, it will completely actuate the process.

Otherwise it will just alert. Like you will get uh a notification sound or something like a buzzer or things like, like maybe it may be uh just maybe a lamp, it will be turned on and turn off something just to alert the driver. So these are the parts of this uh system basically. And uh and in terms of automation, like it can be very low to very high. So the basic thing is like uh no automation which is a common uh vehicle here. So there can be no automation and there can be complete automation. So you can see uh uh like what, how much interference from the human driver is required in this particular figure, like if there is no automation, uh we have to completely control uh even though it is the steering, even though it is the monitoring of driving environment, even though whatever it is, we have to take full control.

And based on the level of automation, like there can be driver assistance. Uh In that case, you'll be getting an assistant for the steering and acceleration and deceleration, things like that only that part will be assisted by this autonomous software. That we'll be using or the A I software that we'll be using. But you can see it is not completely controlled even there, like it is just a driver assistance, it is not automation, it is just assistance. So driver will be there and there will be some software or some features that, that will assist the driver. So some and some driving won't make this uh particular driver assistance system will be taking control, but there will be no full control in that particular system. And there can be the second level which is called partial automation. And in partial automation, you can see like uh when something happens, complete steering control can be taken care by this partially automated vehicle which is a level two vehicle. And then it comes like uh automated driving system monitors the road, like there are two things like human driver monitors the road and automated driving system monitors the road. So in this particular case, which is 34 and five, the automated driving system itself uh will monitor the road.

So in that case, we are having this conditional automation, high automation and full automation like the condition automation is something like we can control the driving environment, we can control the steering and acceleration and deceleration. And uh like all the points we can completely control.

We don't need a driver to interfere in that like we can enable and disable whether we need this complete uh like conditional automation or not. But uh if we are enabling that we don't need a driver anymore uh for the steering movement or anything. So, and there can be this high automation and uh there can be like just a second, I will plug in my battery like in a charger. Ok. Thank you. And then there is this high automation in which we can have the steering acceleration, desolation, monitoring of driver and M and fall back when a software and you can just take control of everything like how the road is, how the traffic is, how the traffic light is and everything will be controlled by this fully automated way.

But it is like a dream for people like uh us like in India, uh we need complete automation there. Uh The Infra Infrastructure may have to like completely change her, like we need a lot of modification to get this fully automated vehicles here. But uh we are having this conditional automation vehicles and partially automated vehicle uh from Toyota and everything. So we are pretty much feeling good here. And so like uh I made this PPT using uh OK. Uh slide carnival and uh like I will share you the CRE credentials if you want like uh if you want this particular thing or anything, just a second. OK. Uh So I got a question from Simon. Uh Are you incorporating research or working with emotional e in this, I asked the post, the work done by companies like effect and smart I OK. So I will come to that point like actually I am working on the research of the testing. Uh like I have to introduce one more short topic. Like I'm actually using a particular software for car maker where I can like I can check the software like whether it is working fine or not, we will be getting a particular software model or the module which is having this feature implemented and we'll be testing that.

So I will be like creating some scenarios like uh using carmaker, which is a particular group. So right now I'm not using emotional A I or not. My company is researching in this emotional a part. It is a separate extreme. So if you want to like know more about that Simony, like I, I can uh get you some information, like I can share you my credentials like uh my linkedin ID or my Instagram ID and we can have a chat in this uh like I can connect you with some like good people in my office so that uh like you will get much information that you required.

OK. OK. Say morning, I'm not sure like whether I'm pronouncing your name, correct? OK. So like I plan for some uh something like uh 15 minutes presentation so that I get five minutes to chat with you people anyway. Uh like I will share my HM I DS here like my linkedin ID and my mail and everything. So if you want to know more like, uh it is not a very small topic. It's a very big topic. Uh Right now I had just 20 minutes, so I had some very small, like small part of that and automation here. So basically we are working mostly on this human machine interface. Like how, how a person can, like if the driver is sitting there and we can even alert the driver that you have to get a copy. Like you seem to be drowsy, you have to get a copy, that kind of alert is coming from the vehicle itself. So like, uh there is a lot of things uh going on with this air as so if you want to know more, like just give me a message, we can, we can chat on that and we can do a lot of chatting, I guess because it is a very, very big topic. Ok, guys. Anyway, thanks for joining. We can like join in the other meetings and other conferences. And uh thank you, sir, I will connect with you. It was a great experience for me.

I hope you guys are enjoying this conference and thank you for home tech network for giving me this opportunity. Thanks a lot of people for joining me.