Rohith Kannan’s Post

If you have a short exit on a busy street, things can get dicey quickly. NoPileups™ Smart Exit tracks cars beyond the exit and stops the conveyor when cars can't leave your lot. Learn more: https://bit.ly/3SlatFx

Data showing Waymos spending a lot of time in parking lots at non-peak times (evenings week-days; 3-9 AM weekends). No surprise they need to park somewhere when demand is low, which counters some of the usual arguments that assume they are on the street 24/7. On the theme of benefits to cities, perhaps it makes sense for robotaxis to partially displace mass transit in a special circumstance as a win/win: - Pricing is the SAME as mass transit (including weekly/monthly passes) - Nobody can be refused a ride who could also ride public transit - Increases safety at off hours by providing individual rides - Supports increasing transit headway (reducing cost of low-occupancy vehicles) + puts idle robotaxis to work In other words, if robotaxis are added to the mass transit system during slack demand, perhaps it is a win win. However, for this to be a good idea it needs to actually be a win/win rather than further degrading mass transit ridership as we saw happen with ride hail during the day. Notes: - Data from a clever Waymo parking lot monitor cam. - Yes, you need down-time for maintenance, but this is an open parking lot so these cars are not getting maintenance other than possibly data transfer - Tiny font horizontal axis is per-hour bins running from 0 to 23 h/t "entertailion" on Reddit https://lnkd.in/efwd6azF

"Adaptive traffic signals" are great! But what if we shifted our objective from increasing the throughput of cars to moving people instead? Because isn't that what transportation is about? i.e., the safe and efficient mobility of "people" and goods? Check out our latest pre-print where we present a novel decentralized adaptive traffic signal control that provides priority to transit and HOV using the max pressure algorithm. https://lnkd.in/e7YUxEZ8

Our Overdrive autonomy stack climbing a steep, eroded, overgrown tank trail in a 12 ton tracked M5. This is real off-road autonomy: no prior maps, no pre-defined path, no teleoperation. Just realtime (1x) video of high-speed point-to-point navigation using onboard sensing and compute.

My recent publication is now available online. This effort emphasizes the differences in driver reaction time across the road network. The reaction time on curves exceeds the standard reaction time by 4% - 12%. This is not the whole story yet it is another step in my journey of understanding driver behavior and how the driver interacts with the surrounding environment. https://lnkd.in/g9Rcmu2j

🔸 Welcome My Friends 💭 🔹 New Embedded project ◼ Project ◾ Traifc Light. ◼ Features. We understand and use the explanation of traffic light colors to regulate the movement of cars and vehicles on roads within cities, especially crowded ones. They are necessary to reduce congestion and prevent car collisions on multi-directional roads, as cars passing at the same time heading to different roads can lead to a complete stop in traffic. ◼ Green Color. The appearance of green means that cars can pass now, and it is also limited to a certain period. ◼ Red Color. It is the color that prevents cars from advancing and allows pedestrians to cross to the other side, Its duration ranges from half a minute to one and a half minutes, after which the signal turns green. ◼ Yallow Color. This color is not approved in all traffic lights, but is only found in some of them, and it means readiness for cars from both directions. ◼ Hardware. 1-LCD 2-LED 3-ATMEGA32 [Github](https://lnkd.in/eZk5Q6hR) #traificlight #embedded #embeddedsystem #embeddedsoftware

Road paint is not "cycling infrastructure".

Reflections on Using Mirrors When Driving. https://hubs.ly/Q02d-Lbd0

Hi. I'm tracking for last day's quite funny story of self-driving cars honking on each other :) The Master Algorithm trolling people :) More details on The Verge website. I hope they start optimizing the "Algorithm" and force the cars to communicate with each other "silently" :) The cars can use TCP/IP, Api, MessageQueue, or a dozen of different protocols instead of horns and sound waves :D PS1: The "management" implemented a mitigation plan (Sic! reorganizing the parking lot for more space) ... PS2: I intentionally avoid using 2 letter acronym related to "Intelligence", because I think it's been abused and lost its real meaning in the last 2 years. The "honking" video => https://lnkd.in/dY8Z9RuQ

At Motional, we embrace a machine learning-first approach, and trajectory prediction is an important part of that ecosystem. Accurately predicting what other objects in the driving environment are going to do is essential to ensuring that our AVs can navigate complex streetscapes in a way that’s safe to the vehicle and everything around it. Learn more: https://lnkd.in/ecSQMv6x