An Open Letter to Lucid: 1,408 Miles In, 3.52 mi/kWh, and a New Owner's Wish List

[knewman]

All great suggestions that have been discussed ad-nauseam here. Make sure you reach out to Lucid with your suggestions, I very much doubt they have anyone on staff who's spending time reading open letters.

I'm not sure if it's a conservative software development budget or an actual product design choice, but Lucid isn't too bothered about providing information to the driver beyond the very basics.

 

[doubletwist]

m not sure if it's a conservative software development budget or an actual product design choice, but Lucid isn't too bothered about providing information to the driver beyond the very basics.

Which never made any sense to me, you'd think a company full of hard core engineers would've wanted as much nerdy data available as possible.

 

[prgarg]

I have sort of given up on having any expectations for any meaningful software update in the near future. Clearly not a priority for Lucid at this point, probably related to the cost savings initiative. Looks like launching 3.0 for Air is Lucid's mic drop moment for this year.

 

[knewman]

Which never made any sense to me, you'd think a company full of hard core engineers would've wanted as much nerdy data available as possible.

I know! I'd love to see notes from early product design meetings to see if there was any dissent from the engineers about the decidedly simple interface we ended up with.

 

[DeaneG]

I know! I'd love to see notes from early product design meetings to see if there was any dissent from the engineers about the decidedly simple interface we ended up with.

There's always a tension between product management and the engineering team. Different goals.

 

[NoahWeb]

What I wanted to share is that I had one of my Grok Bots generate possible coding solutions for Lucid. All assumptions are stated up front (9 total) — Android Automotive OS, vehicle-property access, dual-screen stack, existing profile system, sensor availability, safety constraints, NVIDIA/storage limits on dashcam-type features, etc.

From there it produced draft approaches for each item. Most of them are relatively straightforward once you accept those assumptions: a new efficiency service and HAL view, a one-line color change on the DreamDrive power meter, a user-settable arrival buffer in the router, ambient-light auto-dim, a smaller preconditioning indicator, counting steering-wheel button events toward the hands-on check, tying auto high beams to the driver profile, a seat-symmetry command, and skipping redundant confirmations in the app for low-risk actions.

Nothing is production code, just structured starting points with short Kotlin-style pseudocode. The point was to see how contained some of these changes could be versus how big they feel from the driver’s seat.

While I appreciate @DeaneG take on the friction between the Product Managers and the Engineers....solutions can still be generated and Lucid needs to feel the pressure to implement, test, validate, and push OTA updates. Wish it was Open Source software....

 

[TahoeMichael]

An Open Letter to Lucid: 1,408 Miles In, 3.52 mi/kWh, and a New Owner's Wish List

I just finished my first big drive in my 2023 Air GT: 1,408 miles from California to Texas at 3.52 mi/kWh, with a mere 5 charging stops (not counting overnight rest charges). I came to Lucid after six years and 133,000 miles in a 2017 Tesla Model S 100D, so I know what a mature EV software experience feels like day to day. The Air is a remarkable car! The efficiency, the ride, the engineering are all world-class. That's why I'm writing. The hardware has earned software and UX to match. Here's my list after one long road trip as a new owner:

Driver information and efficiency
1. Add a "HAL" button (Hyper Analytics by Lucid??). It would open a detailed view of how the car is performing: live mi/kWh, trip projections and consumption trends. Put it where the dedicated Themes button sits now, at the bottom of the main screen, and move Themes into a tab under Settings.
2. Make the left main screen more useful too. Your engineers built the most efficient EV on the road, so show it off even outside HAL. Give us live mi/kWh, an efficiency score and more detailed trip info at a glance.
3. Fix the power meter in DreamDrive Pro. Right now it's gray on gray and hard to read. The white regen meter is great. Make the power bar green on a gray background.
4. Add a compass to Nav. What's my heading?

Route planning
5. Route planning is far too conservative. My charging stops came in at 25–35% state of charge (roughly 100–140 miles of range left unused). Let drivers set their own arrival buffer and plan from it.

Displays
6. Auto-dim the screens based on ambient light.
7. Rework the right home screen. Time and temperature show up twice. Use that space for more useful, consolidated data.
8. Prevent burn-in on the left control screen. For example, routinely shift icons about 10 pixels within their existing space.
9. Shrink preconditioning. Drop the giant takeover. Show a small icon at the bottom of the main screen, or a short text label like "Battery Conditioning" above the drive mode (Smooth / Swift / Sprint).
10. Number the volume steps, say 1 to 20.

DreamDrive Pro
11. Count steering-wheel inputs in the hands-on nudge check. Next track, turn signal and volume changes all show an engaged driver.

Media
12. Refresh the Spotify integration. It's clunky.

Comfort and profiles
13. Tie auto high beams to the driver profile, so they stay on or off.
14. Add a seat symmetry button for drivers who don't always have a passenger but want the seats lined up.

Safety and features
15. Build in dashcam recording. I know the NVIDIA architecture may make it hard, but it deserves a real attempt.

Lucid app
16. Stop asking me to confirm actions I already chose. Come on.

None of this takes away from what Lucid built. It's a list from someone who wants the software to match the car. Fellow owners, what would you add?


Admins, being new not sure where to best nest this post/thread.

Seems like great owner feedback. If you're not already aware, you should email it to [email protected].

 

[doubletwist]

What I wanted to share is that I had one of my Grok Bots generate possible coding solutions for Lucid. All assumptions are stated up front (9 total) — Android Automotive OS, vehicle-property access, dual-screen stack, existing profile system, sensor availability, safety constraints, NVIDIA/storage limits on dashcam-type features, etc.

From there it produced draft approaches for each item. Most of them are relatively straightforward once you accept those assumptions: a new efficiency service and HAL view, a one-line color change on the DreamDrive power meter, a user-settable arrival buffer in the router, ambient-light auto-dim, a smaller preconditioning indicator, counting steering-wheel button events toward the hands-on check, tying auto high beams to the driver profile, a seat-symmetry command, and skipping redundant confirmations in the app for low-risk actions.

Nothing is production code, just structured starting points with short Kotlin-style pseudocode. The point was to see how contained some of these changes could be versus how big they feel from the driver’s seat.

While I appreciate @DeaneG take on the friction between the Product Managers and the Engineers....solutions can still be generated and Lucid needs to feel the pressure to implement, test, validate, and push OTA updates. Wish it was Open Source software....

Did you take into the account that AAOS is also running on top of a QNX RTOS (Real Time Operating System)? I believe in a VM.

 

[NoahWeb]

Did you take into the account that AAOS is also running on top of a QNX RTOS (Real Time Operating System)? I believe in a VM.

I honestly gave Grok some general guidance, researched whats available publicly, and let it run with a draft. I’m not a coder by nature.

 

[doubletwist]

Sure but if you just said to help you with AAOS, Grok isn't going to know anything else but what you tell it. And even if you mentioned Lucid , it's a crap shoot whether Grok would realize and take that into account, as there's not all that much detailed information about the underpinnings of Lucid's systems.