The Long Goodbye

[TahoeMichael]

Okay, folks.

I just got a call from Emad Dlala (he was in the office on Mother's Day) that lasted well over an hour. He said he is fine with my posting about the call, so here goes . . . .

Between the 3.4 series of updates and 3.5.1 Lucid switched to a different architectural platform. This was necessary because of technical problems with the old platform (as a Luddite I couldn't quite follow all the ins and outs of this), but that is the root cause of the increase in software issues many owners are seeing. They knew this platform change might tip the tea cart over and it would take a while to put it back upright, but they're working 7 days a week on it. Dlala said these issues include locking/unlocking issues, screen freezes and blackouts, and several other issues Lucid knows owners are experiencing. There is an update coming out later this month and another around July that will address many of the issues. Although these updates will add some new features, they will apparently focus more than many of us thought on reducing existing problems.

He said one of the reasons I am seeming more problems than many other owners is that we have two fully-optioned Dream Editions, and running all those systems taxes the software sometimes beyond its limits. Lucid is working on a change that will take about 30% of the load off the system and give it more headroom to handle things without issue (again, the technical aspects of this were lost on me).

We got into an interesting tangent on the HUD, which Dlala said has been problematic on several fronts. He said that he, like many owners on the forum, has actually turned his off in his car. I told him that one of the most universal complaints on the forum is that the display is too high in the field of vision. He is investigating whether that can be at least partially addressed via software.

Although I'm not particularly exercised about promised features that haven't yet arrived, Dlala knows that many owners are, and he's really trying to pick up the pace of bringing them online. He's particularly interested in getting PAAK in place. However, he said that when he took the job he told Lucid that it was going to take six months for him to take stock of the software work to be done and how best to approach it, and he thinks things are now ready to move forward more quickly.

He assured me Lucid is fully aware of what is hanging in the balance with the software performance of the Cosmos at launch and that the work they're doing now on the Gravity with platform changes, more efficient processing, etc., is setting the stage for that.

I was extremely appreciative of his call and have a bit of restored confidence that we may yet see some light at the end of this tunnel.

I'm an engineer with a lot of experience in complicated development projects where we're building both software and hardware. Everything you relate from what Emad says makes a lot of sense to me, and the timeline he outlines for making improvements seems realistic. I'm baffled that Lucid leadership took so long to realize that software was a crisis for them, but Emad does seem to be taking a sensible approach in the wake of them finally having a big shake-up and putting him in charge.

 

[Need4Sleep]

I'm an engineer with a lot of experience in complicated development projects where we're building both software and hardware. Everything you relate from what Emad says makes a lot of sense to me, and the timeline he outlines for making improvements seems realistic. I'm baffled that Lucid leadership took so long to realize that software was a crisis for them, but Emad does seem to be taking a sensible approach in the wake of them finally having a big shake-up and putting him in charge.

Agreed, and better late than never. With Emad taking the charge on software, Silvio focusing on operational efficiency, and Cory Steuben as director of cost engineering, I think Lucid's leadership is shaping up to take on scale. Hopefully they don't bite off more than they can chew with the midsize launch, and I also hope the learnings on the midsize program (in terms of cost, software stability, cost engineering, etc.) trickle back to the Gravity and Air programs

 

[TahoeMichael]

Agreed, and better late than never. With Emad taking the charge on software, Silvio focusing on operational efficiency, and Cory Steuben as director of cost engineering, I think Lucid's leadership is shaping up to take on scale. Hopefully they don't bite off more than they can chew with the midsize launch, and I also hope the learnings on the midsize program (in terms of cost, software stability, cost engineering, etc.) trickle back to the Gravity and Air programs

I believe they've said that the Cosmos will not be running the exact software platform as the Gravity, but I'd be very surprised if it isn't just the next evolution of the same software base. As such, serious architectural issues in Gravity's software almost certainly impact Cosmos as well. If there's any sanity in the leadership team, they won't let a journalist (or YouTube EV influencer) anywhere the Cosmos software until they're sure they've very solidly dealt with the issues. Lucid has racked up enough of a reputation for software quality problem that everyone in the auto press/influencer space will doubtless be looking for problems the moment they're sitting in one. I believe that Kyle from OOS did an early preview of the iX3 where he commented something to the effect that the software seemed very dialed in. If they don't get a similar reaction when Cosmos is unveiled, they've really screwed up their launch.

On the positive side: the rest of Lucid's engineering is so good that if they can fix their software team, then their vehicles are going to continue to be world leaders.

 

[gaskiller]

I only joined the forum about 3 weeks ago to begin my Gravity research for my next car. And I had already decided to pass. This thread pretty much sums up my decision path in reading through many other threads in this forum...

My test drive gave a great impression. I liked everything about the car except the audio system and exterior colors. The Gravity was #1 on my list. But in reading this, BMW, Cadillac, Polestar, and Rivian forums, I have come to realize that my absolute top priority is that a car's core function must be 100% reliable, like my coffee maker. In my lifetime of using fobs, I have a 0% failure rate. I will not consider a Gravity until I start seeing forum evidence that the car has become 100% reliable from an unlock-start-and-drive perspective. I will get a lesser car in other areas (less luxurious, fewer features, less stylish, etc.) in exchange for basic dependability.

It’s already improved significantly after last update. And there is a posting bias, those with problems post, those without do not. You are depriving yourself of the best SUV ever made.

 

[gaskiller]

One could argue that there is no "perfection" and no one can guarantee100% of the products shipped are good. Go tell that to Lexas/Toyota. Yes, from time-to-time that might be bad parts that escape. But it is not 5%. Even if there were escapes, isn't it the Service Center's job to test out the cars and its accessories (in this case, EOBs) to make sure they don't get to the customers?

BUT: knowingly shipped 5% products that's bad and considers "acceptable" is, (to borrow @hmp10's phrase) absurd!

If you go to the supermarket to buy a box of dozen eggs and there is a high probability that one of the eggs in the carton is rotten, would you go back to the same supermarket?

Let's get back to key FOBs:

> if Lucid KNOWS that about 5% of the key FABs are defective, did Lucid screen them out or ask the supplier to screen them out? If not, why not?

So you expect Lucid to be as good as a manufacturers making cars for almost a century? I like the way you think…….

 

[gaskiller]

That conversation is great news. But it shouldn't have come to a post like this for Lucid's customers to learn a bit more about the reasons they see these issues and what is being done to address them.

Lucid needs to be more open, frankly their reputation is so low, that any information would make it better.

No company is open about issues unless stipulated by SEC reporting guidelines. ……why would you hang your laundry out for everyone to see.

I’ve heard that 100 times….” They should be more open” …….the reality is very different. Why can’t people understand that?

 

[hmp10]

So you expect Lucid to be as good as a manufacturers making cars for almost a century? I like the way you think…….

They're going to have to get pretty damn close, as those manufacturers are who Lucid is competing against. And none of those manufacturers have been doing UI software for anything close to a century. In fact, some haven't been at it any longer than Lucid . . . especially when you take account of the fact that many Lucid engineers and technology employees were at Tesla before joining Lucid at a time Tesla was leading the charge on UI software ahead of many of the legacy manufacturers.

 

[gaskiller]

They're going to have to get pretty damn close, as those manufacturers are who Lucid is competing against. And none of those manufacturers have been doing UI software for anything close to a century. In fact, some haven't been at it any longer than Lucid . . . especially when you take account of the fact that many Lucid engineers and technology employees were at Tesla before joining Lucid at a time Tesla was leading the charge on UI software ahead of many of the legacy manufacturers.

Tesla had a 20 year head start. And many of those other manufacturers you mention, though reliable, the functionality is way behind Lucid. Agree, Lucid made missteps, but these things happen as you develop a whole new stack.

This is the exact reason they came out with the Air and Gravity before their main profitable model. You dial in the software. ICE manufacturers can build a car for today, but not for tomorrow. I bet you those BMW or Mercs don’t get much in software updates and consistent improvements.

 

[gaskiller]

People need to realize, grass is not greener elsewhere…..BMW software not that great

BMW, I am so breaking up with you | TechCrunch

I’m driving my i4 back to the dealership as soon as my lease is over. I'm surprised to say I can't wait.

techcrunch.com

 

[FeelLicks]

Okay, folks.

I just got a call from Emad Dlala (he was in the office on Mother's Day) that lasted well over an hour. He said he is fine with my posting about the call, so here goes . . . .

Between the 3.4 series of updates and 3.5.1 Lucid switched to a different architectural platform. This was necessary because of technical problems with the old platform (as a Luddite I couldn't quite follow all the ins and outs of this), but that is the root cause of the increase in software issues many owners are seeing. They knew this platform change might tip the tea cart over and it would take a while to put it back upright, but they're working 7 days a week on it. Dlala said these issues include locking/unlocking issues, screen freezes and blackouts, and several other issues Lucid knows owners are experiencing. There is an update coming out later this month and another around July that will address many of the issues. Although these updates will add some new features, they will apparently focus more than many of us thought on reducing existing problems.

He said one of the reasons I am seeming more problems than many other owners is that we have two fully-optioned Dream Editions, and running all those systems taxes the software sometimes beyond its limits. Lucid is working on a change that will take about 30% of the load off the system and give it more headroom to handle things without issue (again, the technical aspects of this were lost on me).

We got into an interesting tangent on the HUD, which Dlala said has been problematic on several fronts. He said that he, like many owners on the forum, has actually turned his off in his car. I told him that one of the most universal complaints on the forum is that the display is too high in the field of vision. He is investigating whether that can be at least partially addressed via software.

Although I'm not particularly exercised about promised features that haven't yet arrived, Dlala knows that many owners are, and he's really trying to pick up the pace of bringing them online. He's particularly interested in getting PAAK in place. However, he said that when he took the job he told Lucid that it was going to take six months for him to take stock of the software work to be done and how best to approach it, and he thinks things are now ready to move forward more quickly.

He assured me Lucid is fully aware of what is hanging in the balance with the software performance of the Cosmos at launch and that the work they're doing now on the Gravity with platform changes, more efficient processing, etc., is setting the stage for that.

I was extremely appreciative of his call and have a bit of restored confidence that we may yet see some light at the end of this tunnel.

This is very great news and as a SW engineer sounds very realistic and reasonable. Wish him best of luck and wish Lucid had did this earlier.

 

[drb3333]

People need to realize, grass is not greener elsewhere…..BMW software not that great

BMW, I am so breaking up with you | TechCrunch

I’m driving my i4 back to the dealership as soon as my lease is over. I'm surprised to say I can't wait.

techcrunch.com

Honestly, I am really glad you brought this up. While my car is at a Service Center I have had both a KIA EV and a RAV4 loaner.

We pretty quickly forget just how ordinary these cars actually are compared to our Lucids. Absent my ownership I may never have realized their individual shortcomings.
(I promise you I will never set foot in a POS KIA EV ever again! Just my opinion; YMMV.)

 

[tarator]

I believe they've said that the Cosmos will not be running the exact software platform as the Gravity, but I'd be very surprised if it isn't just the next evolution of the same software base. As such, serious architectural issues in Gravity's software almost certainly impact Cosmos as well. If there's any sanity in the leadership team, they won't let a journalist (or YouTube EV influencer) anywhere the Cosmos software until they're sure they've very solidly dealt with the issues.

If there is any sanity in the leadership team, they won't make the same mistake as with Gravity and create yet another software platform.

 

[hmp10]

Tesla had a 20 year head start. And many of those other manufacturers you mention, though reliable, the functionality is way behind Lucid. Agree, Lucid made missteps, but these things happen as you develop a whole new stack.

This is the exact reason they came out with the Air and Gravity before their main profitable model. You dial in the software. ICE manufacturers can build a car for today, but not for tomorrow. I bet you those BMW or Mercs don’t get much in software updates and consistent improvements.

You're forgetting that Peter Rawlinson, the Chief Engineer of the Lucid Air, was the Chief Engineer of the original Tesla Model S at the time it broke ground with its UI suite of features. That whole effort was under his oversight. And before that, Rawlinson was the head of Advanced Engineering at Lotus, the firm to which Tesla turned for help with the original Roadster because Lucid could not figure out how to handle the power electronics and balance the cells in the battery pack, all part of a battery management system driven by software.

This notion that Lucid launched as a company with zero experience in software while others were light years ahead is just wrong. Take the software management of the powertrain components, the chassis, and other non-UI aspects of the Air. Lucid is not only up to speed with any other manufacturer there; in some areas it is at the front of the pack. If Lucid powertrain software was in the state that its UI software is, not a single Lucid could make it out of a driveway.

The reason Lucid UI software is behind the curve is that they didn't devote the resources and attention to that part of the car, not because Lucid was caught flat-footed by coming late to the party.

 

[borski]

Fair,
With respect to the door unlocking problem, do you think the problem is the FOB or in the car?
I don't like the Lucid FOBs. But I have a hard time seeing the FOB is the issue.

Given that mine is perfect and my wife’s is not, I believe it is, at least in part, the fob.

Of course, I’m not on the inside and don’t know for sure.

 

[BS8899]

Given that mine is perfect and my wife’s is not, I believe it is, at least in part, the fob.

Of course, I’m not on the inside and don’t know for sure.

I am surprised that an astute man like you will make a statement like: "mine is perfect and my wife’s is not". This statement will come back and haunt you in a big way!

Seriously, inasmuch as I like to blame the FOBs, I remain skeptical.

 

[mfdutra]

The reason Lucid UI software is behind the curve is that they didn't devote the resources and attention to that part of the car, not because Lucid was caught flat-footed by coming late to the party.

That's right. It's so much easier to develop software for a dedicated ECU though. Everything is very deterministic.

Developing a lot of different applications that will run concurrently in one bigger computer is a completely different animal. Now you have to deal with myriad of concurrency issues, noisy neighbors, race conditions, I/O all over the place, etc. They really underestimated the task.

 

[Tim-in-CA]

That's right. It's so much easier to develop software for a dedicated ECU though. Everything is very deterministic.

Developing a lot of different applications that will run concurrently in one bigger computer is a completely different animal. Now you have to deal with myriad of concurrency issues, noisy neighbors, race conditions, I/O all over the place, etc. They really underestimated the task.

I believe that this is why Rivian went to the zonal architecture for GEN two of the R1 and launch of R2. This helps improve software development, efficiencies, and hopefully reduce the complexity and issues that arise from having many ECUs to have to deal with. I believe this is the primary reason why Volkswagen did a deal with Rivian for their technology and software.

 

[Gravitas]

Okay, folks.

I just got a call from Emad Dlala (he was in the office on Mother's Day) that lasted well over an hour. He said he is fine with my posting about the call, so here goes . . . .

Between the 3.4 series of updates and 3.5.1 Lucid switched to a different architectural platform. This was necessary because of technical problems with the old platform (as a Luddite I couldn't quite follow all the ins and outs of this), but that is the root cause of the increase in software issues many owners are seeing. They knew this platform change might tip the tea cart over and it would take a while to put it back upright, but they're working 7 days a week on it. Dlala said these issues include locking/unlocking issues, screen freezes and blackouts, and several other issues Lucid knows owners are experiencing. There is an update coming out later this month and another around July that will address many of the issues. Although these updates will add some new features, they will apparently focus more than many of us thought on reducing existing problems.

He said one of the reasons I am seeming more problems than many other owners is that we have two fully-optioned Dream Editions, and running all those systems taxes the software sometimes beyond its limits. Lucid is working on a change that will take about 30% of the load off the system and give it more headroom to handle things without issue (again, the technical aspects of this were lost on me).

We got into an interesting tangent on the HUD, which Dlala said has been problematic on several fronts. He said that he, like many owners on the forum, has actually turned his off in his car. I told him that one of the most universal complaints on the forum is that the display is too high in the field of vision. He is investigating whether that can be at least partially addressed via software.

Although I'm not particularly exercised about promised features that haven't yet arrived, Dlala knows that many owners are, and he's really trying to pick up the pace of bringing them online. He's particularly interested in getting PAAK in place. However, he said that when he took the job he told Lucid that it was going to take six months for him to take stock of the software work to be done and how best to approach it, and he thinks things are now ready to move forward more quickly.

He assured me Lucid is fully aware of what is hanging in the balance with the software performance of the Cosmos at launch and that the work they're doing now on the Gravity with platform changes, more efficient processing, etc., is setting the stage for that.

I was extremely appreciative of his call and have a bit of restored confidence that we may yet see some light at the end of this tunnel.

@hmp10, perhaps a new thread..."The Immediate Return"!

 

[RajD]

Forgive my ignorance:

> if all the FOBs are on the same firmware
>but we say "5% are defective"

Are we saying 5% of the FABs are defective? If so, how are the FOBs defective? Are these "defective FOBs " unable to properly send the signal/codes to the car? If so, are there QA procures that could/should have detected this problem? 5% is a BIG number!

OR

Are we saying, 5 % of the cars are defective vis-a-vis, the FOBs transmit the correct codes, but the defective cars did not actuate the door unlocking?

I'd think Lucid should know the answer to these questions.

My take is, Lucid has been blaming the FOB manufacturers on both the Air and the Gravity and making non-sensible explanation about contract issues with FOB manufacturers that don't make any sense.

I was told by my SC that there was a bad ‘batch’ of fobs that were given to owners and have quality issues with consistent performance. This doesn’t exactly answer your questions but their solution is to replace those fobs. I agree the blame could fall on the fob, the car, or both. And the solution may be the only thing they can do to fix it. But it is being positioned as a fob failure and not the car.

However, there definitely has not been a ‘recall’ of those fobs. It’s almost like they don’t know which owners have which batch of fobs so they have to wait for folks

Honestly, I am really glad you brought this up. While my car is at a Service Center I have had both a KIA EV and a RAV4 loaner.

We pretty quickly forget just how ordinary these cars actually are compared to our Lucids. Absent my ownership I may never have realized their individual shortcomings.
(I promise you I will never set foot in a POS KIA EV ever again! Just my opinion; YMMV.)

this.

We rented a EV9 in Denver snowstorm. That thing was not safe in snow and ice, and it was very bare bones wrt software.

 

[doubletwist]

He said one of the reasons I am seeming more problems than many other owners is that we have two fully-optioned Dream Editions, and running all those systems taxes the software sometimes beyond its limits. Lucid is working on a change that will take about 30% of the load off the system and give it more headroom to handle things without issue (again, the technical aspects of this were lost on me).

At the risk of repeating myself from another thread, this is relevant here too.

If a fully spec'd car was already hitting CPU limits, before even half of their advertised features were deployed and activated, that is a massive failure in their development processes and hardware/software architecture decisions for the Gravity.

They'd better not make the same mistakes for Cosmos/Earth, or it will definitely be the final nail in Lucid's coffin.

For all the shortcomings of Tesla, they have absolutely set the consumer's expectations that the processing power and memory in day 1 cars should have sufficient headroom to handle years of upgrades and additional features.

Hopefully Emad and team are successful in getting the rest of the leadership to understand and act on that for the Cosmos/Earth.