I think TR REALLY needs to push the BETA part a lot more and a lot harder.
AT is an experiment. If you can’t handle being a part of an experiment, then you are not the right person to use it. Not an attack on anyone, I just feel like too many people think this is some special training platform when it isn’t. Maybe someday it will, but it isn’t yet.
We are closer to those people behind the wheel of self driving cars 5 years ago, we are not the guys sitting in the back seat 10 years from now.
Is there a way to use your mobile device in lieu of a head unit for outside workouts?
You can connect and run the regular TR app. Some people did and still do that. It lacks the GPS info and is just the same workout setup as you have inside.
So, not really targeted for or intended for outside use, but without a head unit, its possibly the easiest option.
So you would just load the workout as an inside workout? Im asking because i currently don’t have a head unit.
I still can’t believe people don’t understand this.
I consider myself lucky to not have signed up for the beta, like you I’m under the impression some people confuse testing with early access 
BUT
What @JSTootell said: They should stress the beta status more.
In fact they do the opposite by hyping AT with Nate and the whole crew saying they can’t live anymore without AT because it is soooo awesome. You really can’t blame people if they have expectations at this point …
I know this has been debated over and over again, but the release was maybe ultimately rushed due to the bad publicity the week before …
Don’t get me wrong, I think it is great, that TR is trying to develop something like AT. Since I’m obviously not on AT I can’t judge, but I’m under the impression that most companies would call a product in this phase alpha (or maybe you just hear the complaints and the vast majority is fine?). Don’t confuse this with me saying this openness (is this even a word?) and showing what TR is working on is a bad thing. It’s just … I personally jumped on so many hype trains (yes, I’m middle aged), I can sit this one out 
This is with distance my most wanted missing feature for AT.
If it makes the puzzle a little easier: it would be quite acceptable to me if (the first iteration) considers all unstructured outdoor rides to be “endurance” and only affect that PL.
What I am hoping for is that the machine learning detects that I can do 8 hour endurance rides with ease and decide that a 1 hour endurance TR workout is not going to help much. Or, even better, that I can schedule a 4-8 hour outdoor ride every Sunday and that TR “compensates” by scheduling shorter but more intense workouts during the week.
I very much hope AT / the next generation of TR will be a success. We’ll all benefit.
But the announcement of AT on 25 Feb looked an awful lot like a new product release. The video which accompanied it (nice production values and all) lasts 1 minute and mentions Beta zero times: Adaptive Training: The Right Workout. Every Time. - TrainerRoad
Whether or not the Dylan Johnson YouTube upload on 14 Feb caused Nate and the team to make the release earlier than planned is up for debate. The fact that the beta is still missing such key components as outside workouts in July (including compatibility with Wahoo units and with Strava) suggests it was released too early and promised more than it could initially deliver.
I think there’s general agreement here about the introduction of AT being overly optimistic and looking a lot like early release vs beta. Given that, IMHO there is still a difference between how TR continues to position AT and where it is in development. For example, in last week’s podcast Nate says AT is working outdoors for structured workings on Garmin. What’s left unsaid is (and what I’ve heard from support and read here) is that the survey process is too rigorous on outdoor rides and delivers struggle/fail surveys too often when not actually warranted.
It’s unclear for me if marking the survey as “did not struggle” and then rating it on the easy-all out scale allows for progression or not – my experience would indicate one TR thinks you have failed/struggled you miss out on progressions. Can TR confirm what happens?
Update: just finished Mt Field outdoors, hit midpoint of the three 12-min SS intervals at or slightly above for both AP and NP, and got the struggle survey and no PL updates… and AT recommended drop the 7/14 SS workout by 1.0 PL. From this end-user’s POV outdoor workouts on Garmin are still not working.
In the past, the number of people complaining that TR is too hard far outnumbered those who complained that TR is too easy.
If you think TR might be feeding you workouts that are too easy, Workout Alternates addresses this issue.
Choose a harder workout, complete it successfully, and your new PL in that energy system will match that of the workout. Furthermore, relevant energy systems may also receive a bump.
When TR AT is able to analyze unstructured outdoor in the future, I’d temper my expectations because it’s probably going to be conservative in its estimation, which is probably how it should be.
TRUTH. Its a high priority and being worked on actively. Thanks for your patience, we get the frustration for sure and will provide updates as they’re available.
informed guess…It’s hard to say. Any timeline that I give gets roasted on the forum if we don’t hit it.
I can say that we have people working on it and only it. And that someone is validating the approach right now.
Nate,
Thanks for responding. I fully appreciate how difficult - and risky - it is to put dates on things but also take heart from your update. More power to the elbows of those working on it!
Adrian
Hi @Nate_Pearson I don’t know if this helps at all, but perhaps licensing/buying intervals.icu could help “harvest” the power/exertion data from outdoor rides in a format that TR can then translate and apply to AT? I know TR was focused on taking account of outdoor rides translating into TSS based on HR, but for those of us who ride with power meters outdoors, it seems (to a non-engineer non-tech guy) to be an interesting approach? Thanks for your help.
intervals.icu is great, but TR already have a team of data scientists on staff. i don’t think intervals.icu is doing anything super proprietary or particularly complicated in terms of trying to identify intervals in unstructured rides. it would almost certainly be cheaper and easier to build from scratch than to acquihire and then merge that stuff into the TR codebase
I’m not sure what the benefit would be to AT compared to what we’re already feeding in.
Intervals.icu is a really good tool run by a single guy and / but he’s recently entered into partnership with TrainerDay.
My Wahoo scheduled outdoor rides do not contribute to the adpations 
Correct. Covered elsewhere, some Garmin data is not contributing while others are. Sadly, its not 100% of G use. I haven’t followed the issue closely, but think it has something to do with the way the files go from your head unit to TR. Something about whether it goes via Garmin Connect vs Strava, but I could well be wrong there.
I’m not in the beta, but I rejigged the way my workouts are handed from platform to platform and it is as you say. Garmin to TR works fine every time, Garmin to Strava to TR never works, at least for me…
