Tuesday, September 19, 2017

Missing Animals - Facebook Feature Request

Project Status: Draft


Problem

When pets go missing or when someone finds a lost pet and attempts to track down the owner we often see posts for this on social media such as Facebook.

Solution

I think Facebook should first build the Missing Persons Feature Request I mentioned and then adapt it for use with missing pets. In this way when people in a most likely area the pet might be are scrolling through Facebook they could see a carousel slider of missing pets near them.

This solution is conceptually a bit different from missing persons and for that reason is should not be the same feature but a derivative of it.

Who can post information and updates

While we still need the information to be live and up to date even after it begins propagating around it not longer needs to be tightly controlled for accuracy and details by the authorities. With pets the two main groups controlling the data now need to be the person(s) looking for a missing pet and the person who finds a pet.

Photos - Yes facial recognition for pets works

Often times pet owners will have tons of photos of their them. However, when trying to find a lost pet we normally only see one or two photos shared. I think the user should be able to share a set of many photos. The algorithms can then automatically feature the best "mug shot" like photo of the animal and can build the 360 degree view from the remaining photos in the same way mentioned in the missing persons idea. The rest of the photo group could still be viewed once someone really starts digging into the post.

With more photos it is much easier for a facial recognition to work, in this way when someone starts taking photos of a found animal Facebook can have more information to compare with. So instead of one person posting one photo of a lost dog and one person posting one of a found dog and then hoping that someone across Facebook happens to see both photos and match connect the dots, we can now have Facebook also working hard to do that in the background.

After all the recognition doesn't even have to be very accurate if you think about it. All it really needs to do is find the closest matches within the closest area and suggest those to both parties to let them decide.

Apple Face ID for pets

One cool use for things like Apple Face ID would be to use the facial scanning feature when finding a missing pet (sure, when you are done laughing feel free to continue reading, lol). Think about it though, mapping quickly mapping the animal's face and sharing that rich data set could vastly increase the ability to find a match quicker.

There are two cases where this would come into play. One being if the pet owner was proactive enough to (once this becomes a possibility with Apple Face Id) scan the pet's face before it went missing. If that was the case they could include the scan with the data provided to Facebook. In the other case the person who located an animal could scan and include the information in the Facebook feature.

In either case it is not a requirement but a benefit because either both parties have scanned and the match should be made almost instantly or one of the parties posts a scan and it is more easily matched to the photos posted by the other party.

Accurate and timely information

By building this in as a Facebook feature the company can gear the idea towards making a consistent look and feel with live information that is always up to date and helpful. It will have a clearer photo and a more rich set of information to dig into. It can be shown to the most likely people that need the information and of course like any other post it should be something that can be muted if people don't want to see this in their news feed.

What is a "draft" post?

This post is still in a draft state. Meaning that I have not done any or have done very little research about this topic. It also means that I hope to come back at some point and update the post with clearer descriptions and any applicable graphics related to the clarification of the idea.

I have basically only outlined my rough idea here until a point when I have time to come back and flesh this post out into something more useful to others. I went into a little more detail on another post about why there are so many pipe dream drafts on this blog.

It is notable to mention that the more views this page gets and the more interest it gathers, the more likely I will be to work on it before other drafts on this blog.

Missing Persons - Facebook Feature Request

Project Status: Draft

Problem

When people go missing we often see friends and family start to post on Facebook. This normally happens with good intentions but in an inefficient manner. Also there is little done about follow-up information, meaning if someone has been found the post might continue to be reshared for some time.

Workaround

Currently it is friends and family that we see starting these posts and then trying to personally keep up with comments and interactions directly. 

Solution

I see carousel sliders all the time in Facebook (at least on mobile) suggesting friends I might want to connect with or even marketplace items for sale. I believe missing people to be far more important than selling random items or sponsored ads and posts.

I think that Facebook needs to add a feature which will allow authorities to make official posts of missing persons. These posts should show up as a carousel slider for people in the most likely areas that the person might be found.

Racing the clock

In this way the information may be sent and viewed more quickly by more people in the most effective areas. As done now by friends and family the post slowly gets shared and the locations it propagates to will be all over the map. Imagine if we sent out things like tornado warns like this. People that needed the information would likely not see it or see it way too late. Also many people would see lots of warnings for areas they aren't effected by and soon would starts ignoring each warning or at least taking them much less seriously. Sounds like a pretty bad way of sharing information but sadly that is how we are currently sharing missing persons information on social media.

Accurate and timely information

By building this in as a Facebook feature the company can gear the idea towards making a consistent look and feel with live information that is always up to date and helpful.

So instead of some random Facebook user starting a post and letting it be reshared over and over with stale information, we can now have a card view with live information on it at all times. No need to read through a ton of comments and try to learn if the person has already been found.

Better photos

Every photo on Facebook is processed with computer vision and machine learning to extract faces and learn who people are. This can be used for good in this case because when the authorities publish the missing persons report they can have access to all public photos of the person. The algorithms can automatically pull out the best "mug shot" style recent photo of the person.

In most cases it could probably even automatically edit the last known shirt or top for the person. In addition to the already public Facebook photos to grad from friends and family should still be able to provide authorities with photos for publication. After all, don't we all pretty much wear the same clothes all the time. Meaning for example, if someone went missing and was last know to be wearing that green shirt we have a photo of from two years ago, it should be able to merge that old photo of the shirt onto the most recent photo of a face we are posting.

Why stop there with photos? I would take it a step further for sure. By aggregating several photos of the missing person together Facebook could automatically build a 360 degree view of the person's entire head and shoulders. When you scroll through and first see a group of local current missing persons their may be facing you. Then in the same way you can drag a 360 degree photo or video around you could swipe to view the person from a profile view and so on.

Don't be intrusive

Overwhelming people with information actionable may cause problems. It would not be a good idea to pump too much information out to people, imagine opening Facebook in New York City for example and seeing tons of missing persons listed.

Some people may not want this to show up in their normal news feed at all actually and just like with other posts they should have the option to turn toggle it off. However, I think it should always be a feature you can get to in some way such as when you search for a person's name or the single item makes it to your news feed because one of your direct friends interacted with it. It could also be in the settings as a feature you can turn back on when you want.

The trick is to find the perfect balance of getting the right information to the right people at the right time and not flooding them with so much information that they feel they are unable to help or are completely overwhelmed by the experience.

What is a "draft" post?

This post is still in a draft state. Meaning that I have not done any or have done very little research about this topic. It also means that I hope to come back at some point and update the post with clearer descriptions and any applicable graphics related to the clarification of the idea.

I have basically only outlined my rough idea here until a point when I have time to come back and flesh this post out into something more useful to others. I went into a little more detail on another post about why there are so many pipe dream drafts on this blog.

It is notable to mention that the more views this page gets and the more interest it gathers, the more likely I will be to work on it before other drafts on this blog.

Saturday, September 16, 2017

Locating prison shivs with micriophones

Project Status: Draft


Problem

Prisoners have plenty of time and they far outnumber the guards. Using anything they can find to weaponize they will make what are often referred to as shivs.

Workaround

Currently the best practice is to conduct periodic searches of the cells. Locating and collecting any contraband found. Or in a worse case, finding weapons after they have been used to do harm.

Solution

The idea is to start using technology to assist employees and proactively find weapons made by prisoners. One step in doing this would be to integrate a network of microphones throughout the complex.

The key is that making a shiv out of metal or iron for example will make a distinct sound. The constant rubbing or sawing noises can be picked up by the microphones and then pinpointed to the location in which they were generated.

Already in use for gunshots in some cities

You should check out this quick video showing how a similar system of learning sounds and continuously listening to them is already being used to location shots fired.



Increase safety with more technology not more employees

No extra guards will be hired on to sit around and listen to the audio. Instead, much in the way Shotspotter works, it will provide accurate and timely information to the people already working in the prison.

What is a "draft" post?

This post is still in a draft state. Meaning that I have not done any or have done very little research about this topic. It also means that I hope to come back at some point and update the post with clearer descriptions and any applicable graphics related to the clarification of the idea.

I have basically only outlined my rough idea here until a point when I have time to come back and flesh this post out into something more useful to others. I went into a little more detail on another post about why there are so many pipe dream drafts on this blog.

It is notable to mention that the more views this page gets and the more interest it gathers, the more likely I will be to work on it before other drafts on this blog.

Friday, September 15, 2017

Stab proof clothing for prisoners and guards

Project Status: Draft


Problem

There is a high risk in prison of being stabbed whether you are an inmate or an employee.

Checkout this awesome liquid body armor


The Idea

Once this becomes more feasible to mass produce I think all prisoners and prison staff should have it. Of course it goes without saying that the military needs this for all of their clothing but even as the video shows they are already working on that.

The Problems

The way the material works once treated with this liquid is that any great force, such as those caused by stabbing or projectiles, causes a reaction in which the molecules pull together with a seemingly unbreakable bound.

I think the issue with that is that a the material itself is not impenetrable when faced with slow or gentle forces. Meaning it is possible to cut a material treated with this liquid but not to stab through it quickly with those same scissors.

That means if someone was held down and slowly stabbed this material would provide no defense other than buying time and even a bit of cover from the element of surprise. So while not a perfect solution I would surely want to be wearing it myself if given the option.

What is a "draft" post?

This post is still in a draft state. Meaning that I have not done any or have done very little research about this topic. It also means that I hope to come back at some point and update the post with clearer descriptions and any applicable graphics related to the clarification of the idea.

I have basically only outlined my rough idea here until a point when I have time to come back and flesh this post out into something more useful to others. I went into a little more detail on another post about why there are so many pipe dream drafts on this blog.

It is notable to mention that the more views this page gets and the more interest it gathers, the more likely I will be to work on it before other drafts on this blog.

Thursday, September 14, 2017

Visually interesting loading screens

Project Status: Draft


Problem

Loading screens always seem to be a simple spinning graphic or something of the like, but this concept should almost always be a thing of the past.

Origin of the spinner

It all started back on the first desktop computers, from the mouse icon actually, when the computer was literally doing everything it could to keep up. This was even before the days of the Internet mind you, and the computers froze up some often that a spinning graphic at least let you know that it was still trying to do something. That something normally took quite a while and you couldn't do anything else because the CPU was already maxed out.

Times have changed

Slowly computers got faster and eventually came home Internet. This meant that slowly the bottleneck moved from the limitations of the desktop or laptop and to the speed of the Internet connection itself.

Often meaning that the user could leverage the system and have many webpages open at once trying to download content in the background. In this way, while your were reading the one page that had loaded several others where slowly churning away in the background.

The smartphones

Somehow in moving to smart phones we have retained this concept of the simplistic loading graphics, in much the same way I would say that we retained the image of a floppy disk for the save button.

This is much like buying the most beefed up sports cars out there and not even revving the engine a little bit when you are stuck at a red light. Obviously the light is the bottleneck not the engine, so why can't the driver have a little fun during the wait.

The idea

Now on our smartphones, or whatever devices of choice today, we are no longer limited by the power of the device but the speed of the data getting to it. I say we start revving the engines a little bit.

Instead of watching something spin in a circle, let's watch the data as it gets to the device. Let's start making our screens look a bit more like all of the screens we see in futuristic movies.

For example

Waiting for a movie to buffer? How about we watch those pixels build up or fly in piece by piece instead of watch a few gray dots chase each other around in a circle.

If it's a GIF then let's loop what we have loaded until the entire GIF is loaded and overlay timeline to show our overall progress. In this way we are making sure to visually signal the user that the entire image has not been loaded.

I say that because I did notice Google Plus screwing this up a long time ago. They actually did just that with the GIFs, minus the timeline overlay, which was awesome because you could watch the animations as they loaded. Eventually they dropped the idea however because, like all Google products, they didn't follow through all the way.

They gave the user no indication that the image was still loading. Which resulted in a horrible user experience where you never knew if you had seen the whole animation or if you waited for a bit longer you would start seeing more frames in the loop.

If we are waiting on a webpage to load let's see a graph of the Internet traffic and watch some text scrolling by as things are happening.

Only a temporary fix through 

Ultimately I am trying to say we should have a little fun while at the red light since we are sitting on all this power anyway. After all what is it to waste a tiny bit of gas if we are having a lot of fun in the process.

Don't assume I'm ok with having a red light in the first place though. Eventually those should all go away also, so to speak, and we shouldn't have to ever wait while staring at a loading screen even if it is so awesome we enjoy it.

Overtime the bottleneck will move again to somewhere new and transferring data will not be as slow as it is today, but even in the old days we could leverage the system with multiple windows open at once and today we should be doing the same. No, not the same exactly with multiple windows, but still trying to find clever ways to best utilize the bottleneck.

For example, why are we still often times forcing users to stare at a loading screen in the first place? Why can't we as developers either anticipate the needs of the user to help avoid that stop to some degree or to better respect the user's time once we do need to load something.

There are still tons of times when using a mobile device that I get locked into a screen that I can't back out of because I know I will lose my progress and have to start over again. A better solution would be to allow me to leave the screen, almost any loading screen, and either come back on my own to view the progress or have the system simply alert me when the task is complete.

Slowly the bigger companies are starting to get better at this and eventually the little guys will be able to easily implement it as well, which is exactly why I say that making loading screens more entertaining and attractive to users is awesome but should itself become a thing of the past before too long.

What is a "draft" post?

This post is still in a draft state. Meaning that I have not done any or have done very little research about this topic. It also means that I hope to come back at some point and update the post with clearer descriptions and any applicable graphics related to the clarification of the idea.

I have basically only outlined my rough idea here until a point when I have time to come back and flesh this post out into something more useful to others. I went into a little more detail on another post about why there are so many pipe dream drafts on this blog.

It is notable to mention that the more views this page gets and the more interest it gathers, the more likely I will be to work on it before other drafts on this blog.