A Bouncing Ball To Make Danger Zones Safer | Hacker News Reader

A Bouncing Ball To Make Danger Zones Safer

spectrum.ieee.org

This idea (a throwable ball with imaging and other sensors) and several prototypes have appeared on HN before, the first one around 2 years ago. This exact product has reached the HN front page in the beginning of 2013 [1].

I find it strange that none of the prototypes have actually managed to reach a production stage, given that the potential market - military, LE, emergency responders is (supposedly) there. Maybe the price point doesn't make much sense, given the usual (low) prices for fixed-focus camera modules - there is no need for very high resolution here, and the balls better be semi-disposable.

There are a few other designs (and patents associated with the concept) - [2], [3] and [4].

[1] https://news.ycombinator.com/item?id=4757971

[2] http://www.bostonglobe.com/business/2013/01/21/surveillance-...

[3] http://www.serveball.com/

[4] http://jonaspfeil.de/ballcamera (and HN discussion: https://news.ycombinator.com/item?id=3109899)

My opinion is that it's significantly more difficult for a startup to sell to military/fire departments/etc than to the general public. What could likely happen is that some big company with "good connections" will end up developing a similar product and get the market.

Selling to the government is no easy task.

  What it can’t do quite yet is survive a hard fall, but the company is working on that and expects to have durability testing finished this year...

Isn't its durability kind of one of the major selling points? The primary use of these is to throw them into hazardous environments. It seems to me like getting the electronics into a package that would survive the beating would be priority #1

If they reasonably believe durability is possible (though expensive or otherwise deferrable), it makes perfect sense to build a fragile but feature-impressive initial version, to demonstrate the benefits, then harden later.

Look at their team:

http://bounceimaging.com/?page_id=260

Do you really think they've missed something simple like the right prototyping priorities for selling into their target markets?

That's quite a team. The article's light on details, but I have to assume that they have already been doing durability design and that the article is just light on the details.

If the aim is to gather a panoramic image while in the air, I would have thought priority #1 would be getting a compact sensor that didn't suffer motion blur when flying and spinning at the same time, in low light conditions, with battery power, low weight and a tiny lens.

I mean, my phone camera produces poor images when I'm trying to hold it still, whereas a thrown ball can fly at 80mph or spin at 20-30 revolutions per second.

At SIGGRAPH 2010, Microsoft Research showed off a technique [1] where they attached an IMU (measures acceleration/rotation) to a DSLR. They recorded motion and used it to deblur the image.

A ball flying and spinning through the air should have a fairly simple blur, though of a potentially much higher magnitude. I'm not familiar enough with the state-of-the-art, but my gut tells me that post-facto image processing could yield impressive results.

This also doesn't have to be extremely high-res sensors. A lower-resolution sensor with larger pixels would collect more light, allowing for a faster shutter speed, meaning less blur.

[1]: http://research.microsoft.com/en-us/um/redmond/groups/ivm/im...

I'm not saying it's impossible or anything. If they wanted to they could just put a bright flash on, or wait for the camera to roll to a stop, or buy a crazy high performance sensor.

I'm just saying, if I was running the company, that's the problem I'd be most worried about getting right.

> Isn't its durability kind of one of the major selling points?

Or you make them cheap enough to be semi-disposable. That way it only needs to survive long enough to transmit the images back. If they find a surviving ball that's a bonus.

Sure, but presumably these devices (and included sensors to measure things like temperature, oxygen/CO/CO2 levels, etc) could be much more useful if they would survive being thrown into the building, bouncing around, and coming to rest, and then being able to still provide useful telemetry. They might slowly melt in a fire, but in other scenarios (e.g. a HAZMAT situation), they wouldn't be terribly useful if they couldn't survive long enough to come to rest.

They didn't define "hard fall." It's possible it's fine for throwing about right now, but it wouldn't survive being dropped from a second story window.

  "To Make Danger Zones Safer"

Dangerous for who? Safer from what or whom? This "safety ball" is there to help one group of people shoot at another group of people, and it's wrong to implicitly state that one group is good and the other evil, simply by virtue that one has this Happy Fun Ball, and the other does not. What if both teams have one? Whose safer now?

I bet this would be just as useful in a home invasion, or whatever, but you know we always have to put the "WE'RE SAVING LIVES" spin on every escalation in the Limited Warfare technological arms race.

Do not taunt Happy Fun Ball.

> Dangerous for who? Safer from what or whom?

Firefighters. Collapsing roofs. One example of hundreds.

Things aren't moral. Everything can be used for good or evil. Things aren't something you should rail against. Humans are the problem.

It makes it safer for the one throwing the ball. Apparently you missed or are ignoring that those people are just as likely to be disaster responders.

Sometimes I wonder if I'm on hacker news or "Sky mall".

My favorite part:

> Bounce Imaging is planning to sell an emergency responder version for about $1000, a military version for something under $3000.

Because, you know, military.

Building it with mil-spec parts can legitimately raise the cost. Generally, components have a higher failure rate early in their life, then that failure rate goes down, and eventually goes back up. Mil-spec components are the ones which survive the early failure rate curve, so you're paying for that component, plus the others which didn't survive, plus the initial wear-in process itself.

Now, the $1000 price point on this seems somewhat high, but is difficult to judge without more specific specs on what's in it.

For custom hardware with a substantial fraction of the feature-set of a modern smartphone, plus way more sensors, $1000 seems ambitiously cheap.

They aren't going to sell a billion of them, R&D and tooling costs are going to push the price up a lot.

I know that many fire departments and other first responders are very severely constrained by budgets. There's lots of interesting things out there, but they just don't have budget for gizmos which cost multiple tens of thousands of dollars. A $1000 price point seems much more feasible. Though if you throw this into a burning building, I wonder if it would survive to be used again.

Cameras are crazy cheap, less than a buck. Lenses may be expensive to design, but you might find something off the shelf to do what you need.

The expensive pieces probably go, in rough order: Case and mounting hardware, radio, reasonable quality sensors, embedded processor and support chips, power supply. Don't know where the cameras fit into that, since you want a bunch of them.

If you can't get a COGS under $200 you don't deserve to be in business.

Actually they could cost the same. It just said under 3000.

I'd assume that these would end up being re-usable in many circumstances, though certainly not all.

It sounds like they still need to figure out the durability issue -- but at that point it could be re-usable in some cases I'd think (does seem like they'd figure out durability first though).

Generally it goes Consumer, Industrial, Automotive, Medical and then Military Grade when it comes to chips. Differences can include higher heat tolerances or additional fail safes, such as additional watch dog timers, etc.

Would not a small hardened quadrocoptor carrying sensors be better - presumably you could develop software to autonomously navigate inside a building.

Not until a quadrocopter can deal with a shotgun blast, buddy.

Reply on news.ycombinator.com