Sponsored
OpenPilot on the F-150 (BlueCruise alternative)
- Thread starter hitchhiker
- Start date
- Watchers 92
Spiffy
Well-known member
Level 2 autopilot? I understand open pilot is level 2 fir the f150. Are there different capabilities based on different vehicles?
Thankfully there are really good people forking the software to bring the f150 in.
Will openpilot have the capabilities of advancing into level 3?
What additional mechanical or electrical hardware are required to do full or close to full autonomous driving?
I have suspicions that Ford has stopped or slowed development of blucruse. 1.2 is far from being considered a stable platform. 1.3 would have to be really good before anyone is going to find $800.00 value. Especially when the comma3/openpilot is so good.
I ordered the comma3. It's more than the 800.00 ford will charge. I like the integration that blucruse provides but that's kind of pointless if I can't keep it engaged for more than 3 miles and the constant keep hands on wheels or false eye nags.
If there is a rapidly advancing platform like openpilot why spend millions like ford and others do on R&D? Just sit around and see what the public comes up with?
From what I have gathered OP is kicking ass. Then you consider fully atonimous cars. What systems are being use there?
Thankfully there are really good people forking the software to bring the f150 in.
Will openpilot have the capabilities of advancing into level 3?
What additional mechanical or electrical hardware are required to do full or close to full autonomous driving?
I have suspicions that Ford has stopped or slowed development of blucruse. 1.2 is far from being considered a stable platform. 1.3 would have to be really good before anyone is going to find $800.00 value. Especially when the comma3/openpilot is so good.
I ordered the comma3. It's more than the 800.00 ford will charge. I like the integration that blucruse provides but that's kind of pointless if I can't keep it engaged for more than 3 miles and the constant keep hands on wheels or false eye nags.
If there is a rapidly advancing platform like openpilot why spend millions like ford and others do on R&D? Just sit around and see what the public comes up with?
From what I have gathered OP is kicking ass. Then you consider fully atonimous cars. What systems are being use there?
Last edited:
labjr1
Well-known member
This link will explain the journey to level 3 which is Open Pilot's intent: to get to level 3 (driver does not have to pay attention on certain sections of road) you have to be able to prove through data that your self driving system is consistently safer than humans on that stretch of road and you have to be able to convince insurance companies through the data to cover cars without human intervention or attention. That is the goal and serious progress is made every week. Yes a Comma3x is more expensive than 1 year of subscription to BlueCruise - but the Comma3x doesn't cost you anything in year 2,3,4... and it is dramatically better than BlueCruise in that it works on most roads (marked or unmarked) and not just limited access highways. Since I have been using my OP (3 months) I have seen many new features added in Sunny Pilot experimental mode including lane change with BLIS with turn signal, stop light recognition, visual and map recognition of sharp curves with automatic slow down, longitudinal control of the vehicle by OP,..., features seem to be added weekly.
Sponsored
Spiffy
Well-known member
Can you explain the development path
To me it looks like open pilot puts its muscle on just a narrow line of cars. Honda, Nissan.
Then because the project is open source that's where the forks come in. Like the f150. The developers are not concerned with us 150 owners for now but the group involvement brings this?
Forks are developed, tested then published after they are safe? There are always forks that have capabilitys greater than what's made public?
I'm just learning about how open source software is developed. Please correct me.
labjr1
Well-known member
OP wants to produce code for as many vehicles as possible. Many auto manufacturers try to prevent others from reading their signaling so it takes extensive testing to get the information needed to make the systems work together. They originally focused on those brands that were easy to work with and have relied on open source developers to expand the product lines supported. Fortunately adding a new brand is not like writing new code. It is translating the manufacturers signaling information into something OP can understand and use with their existing code.
HammaMan
Well-known member
Comma won't call it L3 for a while at best (if ever). It limits liability. It's more than capable of doing door-to-door without driver intervention in some iterations. The only competitor there is currently is tesla's ill-named FSD which is still a $12k option and is married to the vehicle it's applied to. In its current form it bests all ADAS systems for the most part. It's still got a long way to go before it's a butler of sorts, but the fact it's making auto makers look foolish and overpriced is amazing. You buy the hardware, the software is free. I don't see how car companies are going to compete outside of buying them and shutting it down, which I don't see being an option. They're moving towards encryption (car companies) which is a few years off yet.
Ajzride
Well-known member
2024 Mustang is already encrypted between the IPMA and PSCM, I'm hoping 2024 F150 doesn't follow suit.
labjr1
Well-known member
Comma relieved themselves of the majority of liability claims thanks to the separation of the hardware from the Open Source software. They sell a hardware device that can act as a video recorder as well as run other software. They were forced to by the NHTSA in order to stay in business. Since they don't develop the Open Pilot code or any of its branches and since the software does not come loaded on any of its devices it makes it hard to hold them accountable.
Sponsored
labjr1
Well-known member
I would hope that some manufacturers recognize that they either need to provide a competitive product at a competitive price or provide an open platform for 3rd party development like OP. Ford's current model of charging an expensive annual subscription while providing an inferior product is not a good long term marketing strategy. If you are locked into the Ford solution due to the use of encryption in the data stream and their product is not as good or is dramatically more expensive people will revolt.. I applaud the use of encrypting the vehicle control data to avoid outside interference by hackers, but they need to provide the ability for 3rd parties to use signed encryption keys to work with their solution. If not they will lose customers in a big way. Many of us love the self driving tech and purchased our vehicles with that capability in mind. (I cancelled my order for a Raptor because they were not supporting BlueCruise in it).
superfordman
Well-known member
Yup, but unfortunately if I had to put money on it, we know it will, the big question will just be when. Hopefully since nothing much interior wise is changing on the 24s they will be spared. However I would almost bank on the next refresh/interior refresh with the new fancy multi-screen APIM will implement this for sure.
I would love to believe that, but I would imagine Ford wants it their way or the highway. I imagine outside malicious sources are target number one, but if they can keep anyone out from using their system better than they can in the process, why not. Hopefully someone will crack the code on it eventually, but will be curious to see how it rolls out and what impacts it has.
Ajzride
Well-known member
PSA: YOU MUST UNPLUG YOUR COMMA BEFORE USING FORSCAN OR FDRS
@hitchhiker can you please add that to the first post?
@hitchhiker can you please add that to the first post?
labjr1
Well-known member
Ajzride is spot on - Make sure you unplug the USB power from the Comma3x when updating via FDRS or Forscan. The Comma3x will cause issues if it is powered on during updates. I have first hand knowledge ?
Spiffy
Well-known member
Open Pilot is similar to the development of aircraft autopilot systems used in experimental planes. In those cases, the builders installed everything themselves, absolving the main company from liability. Interestingly, experimental planes had a crash rate of less than 1 to 1, making it easier to obtain insurance coverage compared to certified planes like Cessna or Beechcraft.
The systems I was involved in were more advanced than what certified companies offered at 50 times the cost. Liability issues often burdened these companies, forcing them to take a conservative approach.
When the company I worked for began outperforming the larger, established companies, one of them decided to acquire us.
It would be quite difficult to win judgement against comma. No way to go after the public code. Maybe whomever published a certain fork but that would be a stretch.
This whole project is exciting.
Don't be surprised if openpilot and comma get bought up. They are no different than a independent drug research company trying to develop something that eventually gets picked up by Pfizer for millions or billions.
My question is. Is the current hardware capable of performing equal to what Tesla is proposing to do with their beta version?
Can the crowd modifications of the public source code be made to perform on the current version of comma to achieve level 3. What ever that actually is.
I haven't been this excited in something like this in years.
This could be the perfect storm in advancement of automation. Especially with public source involvement.
The systems I was involved in were more advanced than what certified companies offered at 50 times the cost. Liability issues often burdened these companies, forcing them to take a conservative approach.
When the company I worked for began outperforming the larger, established companies, one of them decided to acquire us.
It would be quite difficult to win judgement against comma. No way to go after the public code. Maybe whomever published a certain fork but that would be a stretch.
This whole project is exciting.
Don't be surprised if openpilot and comma get bought up. They are no different than a independent drug research company trying to develop something that eventually gets picked up by Pfizer for millions or billions.
My question is. Is the current hardware capable of performing equal to what Tesla is proposing to do with their beta version?
Can the crowd modifications of the public source code be made to perform on the current version of comma to achieve level 3. What ever that actually is.
I haven't been this excited in something like this in years.
This could be the perfect storm in advancement of automation. Especially with public source involvement.
Sponsored
