Robot Not a Dev? Buy Now
Not a Dev?

Misty Community Forum

Initial positive and negative comments from people receiving Anki's Vector

People have been receiving Vector today. A lot of the comments are not at all positive.

This could be a useful road-map of issues that Misty’s roll-out could strive to avoid.


There are people losing their shit in those comments because Anki did not pre-ship their robots to ‘warehouses’ around the country like Apple did with iPhone XS so they would be able to deliver to everyone on the same day… talk about ridiculous expectations. I swear, every time I read comments I lose even more respect for humanity.

Come on robot uprising!

This comment illustrates a need for properly calibrating expectations around vision/mapping/voice especially for such a small device as Vector. As someone who has worked with CV stuff for a long time not only is the above not surprising but it was honestly entirely expected and I think they’ve done remarkably well given the hardware.


Agreed @Vlad. But, I like to look at the positives :slight_smile:

1 Like

Cherry picking some personality feedback:
" Only cute the first time."
“Give us different options and color schemes. Male, female, snooty personality or comedic… I think that’s what we are on our way towards…”

1 Like

From one of the comments:

It’s an understatement to call that an understatement :stuck_out_tongue:


You can change the eye color :wink:

another critical comment from many users is operational duration on battery; e.g.,

Charging; it seems like everyone’s complaint about the battery life. I can accept that he uses quite a bit of battery power. I think it would be helpful to have some sort of signal that he is starting to run low or needs to go home. That way I won’t keep trying to get him to do something when he’s just trying to get to this charger. And also if I’ve moved him, that I know to get him back within range of his charger.


Having played with Vector for 2 days and read comments in the Kickstarter Anki Forum I’m fortunate that I haven’t had most of the issues described.
I definitely agree with you that battery life is an issue.
Also I agree with others posts that facial recognition is an issue. I wear glasses, which seems to be a known issue. But since I always am wearing glasses I don’t see much point in taking them off. After taking a picture of me without glasses, however, Vector correctly identified me with my glasses on.
And here is my most serious issue. I placed Vector on a roomy table with a flat edge. He has fallen off into my hand at least 8 times in 20 minutes. The first time he did this, to catch him I had to let my iPad Pro fall to a tile floor. I figured the iPad was a lot less likely to be damaged.
All in all Anki has work to do. We’ll see if the update tomorrow helps on these issues.

So I think we can all agree that on reading the comments on the Kickstarter Anki Vector forum that Vector’s initial introduction to users has been mixed. Two questions come to mind.

First question: assuming that Anki knew about these issues (and it’s hard to see how they couldn’t know if they did any reasonable initial testing) would they have been better off delaying the rollout. If the major update scheduled for Friday fixes these issues, they could have chosen to rollout Vector a few days until the update was available.

Second question: are there lessons here for the rollout of Misty II. If so, what would you say they are?

1 Like

They clearly expected some things to be suboptimal and they attempted to communicate this in the last update prior to shipping. In that they called out the following:

That said, we’re already planning updates for Vector based on behavior we’re seeing. There are a lot of different systems working together inside Vector, and while the initial balance is okay, we’re tweaking a lot of factors that will make him even better. The update will automatically download to your robot when released on Friday, Oct 12 - our official Vector retail release. Here are a few of the things we’re working on:

  • Eye Contact: Right now, Vector’s not making as much eye contact as we want him to, especially while exploring the area around him. That’s a weird thing to type, and maybe not as technical as you’d expect, but eye contact is an important part of how we bring him to life. We want Vector to boldly explore his surroundings, but still turn and look for you, just to make sure you’re watching.
  • More Socializing: After Vector meets you, he doesn’t greet you as much as we want him to, and we want him to demonstrate his familiarity with you by saying your name more. We’re also tweaking how often he requests interactions, based on how we see him perform with new owners like you, because the last thing you want is a needy robot.
  • Exploration and Activity: We’re actively tweaking when Vector should be curious enough to explore his environment, versus when he’s sitting still. It’s important to find the right balance, and we’re adjusting a lot of factors to dial this in.

I think they were trying to be upfront and balanced the idea of shipping late with more features and shipping on time with fewer and determined the latter was better.

It is not a bad idea, It is actually quite great to have a product that improves week over week versus is the same as when you received it. I am bummed because the SDK isn’t going to be released until next year, right? That is the biggest miss for me.

As for Misty, we are struggling with our own version of this as we look to production timelines, feature sets, and customer expectations. We’ve got our work cut out for us int he next few weeks on these fronts.

1 Like

Not sure I’d really agree with this. It’s hard to base your opinion on a bunch of disgruntled messages on Kickstarter comments. Honestly, from my point of view their rollout has gone pretty smoothly for the tech industry. Much larger companies have had much rockier rollouts.

I’m sure some issues were known while others were not. You can’t possibly test for all edge cases, no matter how big your QA department is. Case and point, there are so many wifi & bluetooth implementations out there and they’re all slightly quirky. It’s almost a given you will hit some devices that don’t behave as you’d like.

You’re also in a no-win situation at that point. Had they delayed a week, there would have been probably a bigger portion of their customers who would have lambasted them for that. And inevitably their new patch won’t fix everything. If you wait until stuff is perfect you’ll never ship. Unless you’re Apple and literally have hundreds of billions in the bank it’s hard to do that (and even Apple ships with tons of problems).

Personally, I’d say the lesson is to do a better job of managing expectations and communicating continuously with our users. Not everyone will be satisfied by this, but I know I will cut a product a LOT more slack if there is great communication of issues, expectations, and planned fixes. The other lesson I’d say is to not target consumer market any time soon :stuck_out_tongue:


I totally agree with your two take-always, @Vlad. Thanks to both you and @Ben for responding.