The new forums will be named Coin Return (based on the most recent vote)! You can check on the status and timeline of the transition to the new forums here.
Please vote in the Forum Structure Poll. Polling will close at 2PM EST on January 21, 2025.
Dropcam aquired by Nest
Thomamelas
Only one man can kill this many Russians. Bring his guitar to me! Registered User regular
Only one man can kill this many Russians. Bring his guitar to me! Registered User regular
So late, late this afternoon, Nest announced they were acquiring Dropcam for $555 million. So who is Dropcam, and why would Nest want them? Why would these forums care?
Dropcam produces IP based surveillance cameras with a hosted cloud solution for storage. You buy the camera, point it where it needs to be, set it up and then you can store your video offsite at Dropcam's cloud at AWS. Their big differentiators are the ease of set up combined with the hosted storage along with a very low price point. Their target market tends to be the 1 to 4 camera installs, usually in the home. They aren't the only player in that space but they are the ones closest to being successful. As for why Nest would want them, it's likely a response to Honeywell. Earlier this week Honeywell announced their new Lyric thermostat intended to compete directly with Nest's learning thermostat. But in the larger picture it's likely that Nest is trying to set up a wi-fi based automated home. They wouldn't be the first wireless automated home option but they could likely be the easiest.
Nest has said that they completely financed the purchase of Dropcam without Google. But some people are likely to have concerns about Google, stored video in the cloud and privacy. The speculation is that this was likely announced at the end of the day on Friday to use a common technique to bury the announcement for that very reason.
Q. How big is Dropcam?
A. Most of the estimates I've seen put them $30 to $60 million a year in revenue. That works out to a few hundred thousand cameras per year along with 40% of Dropcam users paying a monthly fee for video storage. They may be 2% of the total surveillance cameras sold but they are likely a fraction of that in terms revenue numbers. Over all they are a pretty small player but they are in a position to grow. Their main challenger in that space has been Axis with their Axis Hosted Video Solution. But their costs for it have been 3 to 4 times as much as Dropcam's and it has to go through resellers. So they potentially have a market to themselves. But it's a fairly limited one.
Q. Oh god, Google is gonna watch everything I do!
A. Dropcam sends it's stream back to the hosted side to run motion detection and people detection. So there is a constant stream going back to them. In theory there is nothing stopping them from recording. But the reality is that it's not likely to produce something Google would be interested in but you can produce staggeringly massive amount of data an hour. I've worked on systems producing terabytes an hour and it's likely Dropcam would be producing even more data an hour. Storage is cheap but at that scale it's still a hell of a lot of storage. The trick here is how would Google monetize the video data? Analytics are out. Most of them won't produce the kinda data that marketers would find useful. And they tend to be super dependent on camera placement. Seriously, getting them dialed in is a massive pain in the ass.
Q. What else does Dropcam bring to the table?
A. Well they just announced their Tabs product. They are accelerometer based sensors you place on a door. They aren't quite replacements for a home security system but are more for figuring out who is home and who left. They don't really replicate glassbreak sensors but you could use them as a form of geofencing.
Q. Did Nest over pay?
A. Well it's a 10x revenue multiple. Which isn't abnormal for a Silcon Valley tech start up. But it's twice what you'd expect for a company in the surveillance sector.
Dropcam produces IP based surveillance cameras with a hosted cloud solution for storage. You buy the camera, point it where it needs to be, set it up and then you can store your video offsite at Dropcam's cloud at AWS. Their big differentiators are the ease of set up combined with the hosted storage along with a very low price point. Their target market tends to be the 1 to 4 camera installs, usually in the home. They aren't the only player in that space but they are the ones closest to being successful. As for why Nest would want them, it's likely a response to Honeywell. Earlier this week Honeywell announced their new Lyric thermostat intended to compete directly with Nest's learning thermostat. But in the larger picture it's likely that Nest is trying to set up a wi-fi based automated home. They wouldn't be the first wireless automated home option but they could likely be the easiest.
Nest has said that they completely financed the purchase of Dropcam without Google. But some people are likely to have concerns about Google, stored video in the cloud and privacy. The speculation is that this was likely announced at the end of the day on Friday to use a common technique to bury the announcement for that very reason.
Q. How big is Dropcam?
A. Most of the estimates I've seen put them $30 to $60 million a year in revenue. That works out to a few hundred thousand cameras per year along with 40% of Dropcam users paying a monthly fee for video storage. They may be 2% of the total surveillance cameras sold but they are likely a fraction of that in terms revenue numbers. Over all they are a pretty small player but they are in a position to grow. Their main challenger in that space has been Axis with their Axis Hosted Video Solution. But their costs for it have been 3 to 4 times as much as Dropcam's and it has to go through resellers. So they potentially have a market to themselves. But it's a fairly limited one.
Q. Oh god, Google is gonna watch everything I do!
A. Dropcam sends it's stream back to the hosted side to run motion detection and people detection. So there is a constant stream going back to them. In theory there is nothing stopping them from recording. But the reality is that it's not likely to produce something Google would be interested in but you can produce staggeringly massive amount of data an hour. I've worked on systems producing terabytes an hour and it's likely Dropcam would be producing even more data an hour. Storage is cheap but at that scale it's still a hell of a lot of storage. The trick here is how would Google monetize the video data? Analytics are out. Most of them won't produce the kinda data that marketers would find useful. And they tend to be super dependent on camera placement. Seriously, getting them dialed in is a massive pain in the ass.
Q. What else does Dropcam bring to the table?
A. Well they just announced their Tabs product. They are accelerometer based sensors you place on a door. They aren't quite replacements for a home security system but are more for figuring out who is home and who left. They don't really replicate glassbreak sensors but you could use them as a form of geofencing.
Q. Did Nest over pay?
A. Well it's a 10x revenue multiple. Which isn't abnormal for a Silcon Valley tech start up. But it's twice what you'd expect for a company in the surveillance sector.
+3
Posts
As I understand it the bulk of Dropcam's cameras are deployed inside. I haven't taken one apart but they don't look particular weather resistant. Plus you get a lot of people using them to monitor pets/babys.
I'm sure the ISPs will love that. Might be another Net Neutrality issue too.
In theory, you could work up, heights, ages, races, body types, brands of products used, and income levels, probably identify any TV show being watched. Some of this... would be a challenge... admittedly.
If person detection is occurring at the server level, time stamps of when interesting events occur are already being created, and they have access to unencrypted copies of the streams. At a large data center it might be cost effective to run asynchronous person identification processes against it. It would basically only cost electricity durning off peak hours. So, it would need to be great data.
While I don't think any of this is very likely, or very dangerous, motion detection in home is very simple and would cut down on bandwidth utilization. Upload that to a cloud encrypted store in delayed bursts, and you get a product... I would consider if I was actually looking for this sort of thing.
Thanks for the heads up!
The only person I know with a Dropcam, a co-worker at my last job, had an in-home setup. He was an older gentlemen, with a family, had it setup in the living room and would get alerts if there was motion detected during the daytime (I guess when everyone is supposed to be at work/school). I was by his machine once when he got a notification, and then he checked it and his son was playing video games or something. I think it was recommended to him also by someone who had the same setup, but this is obviously a small sample size.
The problem comes from distinguishing between a human and say a dog. It's not impossible, most analytics packages have calibration for this kind of thing. But they tend to be incredibly finicky and you're still likely get a lot of false positives and negatives. That's not impossible for Dropcam to do on their side but there is one factor they can't deal with. Camera placement. To get the analytics to work anywhere near tolerable you have to spend a lot of time thinking about lighting and angles. So the analytics will produce so much junk data it's not even funny.
Almost all of what you proposed isn't a challenge. It's an impossible task with the current state of analytics. You're proposing things that are only Hollywood tech for now.
I have a camera inside my house and my girlfriend just helped set up a camera at a coworker's house on like Thursday.
Like Thom said, your typical IP camera (from like Dropcam, Foscam, D-Link, etc) isn't terribly weather resistant. They make outdoor varieties too.
the "no true scotch man" fallacy.
For DIY stuff yes. Three reasons generally. The form factors of cube and piece of shit PTZ don't lend themselves to outdoor rated designs. Also it keeps the cost down significantly. And most DIYers don't want to run cable through a wall. D-Link is kinda weird in there because they have a consumer line and a professional line. It's not awful but they just by the cameras from a so-so OEM.
As an aside, the only use for a Foscam camera is to occupy space. Such pieces of dog shit.
https://developer.nest.com/
http://arstechnica.com/gadgets/2014/06/ok-google-crank-the-ac-nest-announces-new-smart-home-api/