Axis Partner Elder Care Video Analytics (Smartervision)

By Brian Karas, Published Dec 07, 2016, 04:10am EST

Can video analytics be used to improve the care of the elderly?

Axis and a video analytics startup, Smartervision, are working together to do so. In this note, we examine the solution, its pricing, objectives and potential limitations.

Analytics *** ***** ****

*** ************* ******* ** targeted ** ********** ** elderly ****** ****** * room, **** ** ** an ********-****** ********, ** within * ******* *********. The **** ** ** detect ********* **** * caregiver ***** ** ** automatically *******, **** ** when *** ****** *****, or ***** ** ****** to *** **** ****** a ********** ****** ** time ***** ***** **** a ******** ** ***** area. **** *** ****** the ********* **** ********** **** to check ** ********' ****-*****, which ***** ** *** caregivers *** ***** *****, and *** **** *** patient **** ******* *** independence.

** ******* ************ *** system ** **** ** care **********, *** ****** are **** ** **-**** staff, *** **** ***** also ** **** ** off-site *******, **** ** family *******.

*************'* ********* ***** ***** some ******** ** *** alerts *** ********* **** of *** *******:

Standard ********** ******

********* ** *************, *** typical ********** ** * single ****** **** * ~100 ****** **** ***** to ***** * ******-***** room ******. *********, *** camera ** ********** ** observer *****/**** ***** (** the **** ******, ** to ** ******** ********), and *** ******** **** of *** ****** ***** monitored.

* ****** **** *** camera ******** ** ***** below:

 

Analyitcs *************

*** ** **** ***** can ** ******* ** the **** (***** ** red *** ***** ** the ***** *****), *** then ***** *** ** applied ** ***** *****.

**** ** *** ***** are *****-*****, **** ** to ********* ** * person *** *********** ****** and ******* ********** ** the ******** ***** *** of ****** *** * timer ** ******* **** a ****** ** **** entering *** ********. ** an ***** **** *** bathroom ** *** ******** within * ****** **** limit, ** ***** *** be **** ** * *********.

******** ***** ** ****** analytics ***** **** ** elder ****:

  • ****** ************ (********* ******) in ***
  • ****** ****** ******* **** through **** **** ** night
  • ****** **** * ****** gets *** ** *** at ***** 
  • ****** **** * ****** has **** ** **** and *** ******** ** set ****
  • ****** ***** (********* ** beta)

Camera ** ******-***** *******

*** ********* *** ** camera-based *** **** ****-******* cameras, ** ******-*****, *** the ******* **** ******-***** analytics *** ******** ** get **** *************, *** this ** *** **** common ***********. ** ******* given *** **-** ******* supported ** * ****** "mid-range" (***********) ******.

Camera ************

*** ******* **** ***** analytics ******* * ******* 12fps ****** ** ******* resolution, *** *** ****** higher *********** ** ****. Ideally *** ****** ****** have ******* ** ** see **** ******** ** night. *** **** ******* they ********* ********-**** *** **** ******, but **** ********* ********-*********-**** ************. 

Privacy ********** / *****

****** ****** *** ******** with * ****-******* ******* to ******* * ***** of ******* *** *** person ***** *********, ****** this ******** *** ******-***** scenario.  ************, *** ****** does *** ****** *****, it ** ******** ** be **** ** ** event ******, ****** ** does ***** **** ********** to ** *******.

***** ********** *** ** ************ is *******. *** ******* says **** **** *** looked **** **** *** and *** *** **** of *** *********.

System ***********

*** ********* ****** ******** a ******, ** ************* between *** ****** ******* differently. **** ***** ** mostly ******* ** ***** with **** * ****** occupant, *** *** **** cause ******** ** ********** or ******** *** ** the ****, ** *** system ****** **** ** a ****** **** ******* the **** *** * patient ** *********.

************* ****** *** ***** alarm **** ** ******* in ********** *** **** 20-30% (******* **** *** of ***** *** ******, 20-30 **** ** ***** alarms). *** ******* ***** was * ****** ******** with * ** **** on ********** ******** ****** from *** ** ******* light ** *** ****.

*******

***** ** * **** license ** ~$***/****** *** the **** ******** *** one ********* ***** (**** functions, ******** *********, ** Bed *********), *** **** an ********** ******* *** of ~$** *** **** additional ********* *****. ***** prices *** ****, *** dealers *** ******** *********. Optional *** *** ******** and ******* *** **% of ******** *****.

* ****-***** ******* ***** is **** ********* *** ~$21/mo. *** * *-**** ********, which ***** *** ***** cost ** ********* ******** higher (~$*** *** * years ** **** ***** vs ~$*** *** ******** software *********), *** ******* up-front ******* *****.

** ******** ** *** software *******, ***** **** need * ****** *** typically * ****** ** order ** *** *** software, ********** *** ******* cost.

Analytics ** ***** **** *******

**** **** ********** **** already **** **** **** of * ***** **** system, *********** **** * wireless ***, ***** ****** residents ** ****** *** assistance ** ******. ***** analytics-based ******* *** *********** detect ********* ***** *** patient ** ****** ** get ** *** ***** call ******, ** **** not **** * ****** fob. *** ********* ***** likely ** ** ******** to ***** **** *********, rather **** ** ******** alternative.

Potential *** ********* *********

**** **** * **% false ***** ****, **** solution ***** **** **** benefit ** ********** **** need ** ******** ***** time. *** ******* *** broken **** **** * 3 **** ****** ***** be ******* ********** ** a *** ***** ** pay *** * ********* at * ********. ** long ** *** ***** alarms **** *** ****-********* to ****** *** ******* the *** ****** ***** be * **** *** safety *******, ******** ********** to ** ******* ** residents ** **** ** assistance.

* *** ***** ** ***** video ********* *** ***** care *** ** ******* concerns. **** ** *** cameras ** *** ****** video, *** ********* ****** of *** *******, **-**** cameras ***** ** **** as ******** ** ******** who ** *** ***** *** the ********** ***** *** are ********* ***** ***** privacy.

Comments (6)

Undisclosed #1

IPVMU Certified | 12/07/16 03:07pm

Additionally, the system does not record video, it is intended to be used as an event sensor, though it does allow live monitoring to be enabled.

I don't think I've ever heard of an event system where you can't review that data that caused the event.

So, when it triggers on a person because a person has 'fallen', and the caretaker is not looking at the monitor at that exact moment, what do they do, just see if they look ok now and/or ask them if they fell?

Wouldn't that get old quickly with a false rate so high?

Avatar

Brian Karas

IPVM | In reply to Undisclosed #1 | 12/07/16 04:01pm

I don't think I've ever heard of an event system where you can't review that data that caused the event.

Sure you have. Burglar alarms, people counters using beams or motion detectors, and so forth. Lots of event systems do not have video data to go with the event, and really only tell you a sensor triggered, but do not give any details of how/why the trigger happened.

So, when it triggers on a person because a person has 'fallen', and the caretaker is not looking at the monitor at that exact moment, what do they do, just see if they look ok now and/or ask them if they fell?

I do not think the person is really meant to be looking at a monitor anyway. Think of it more like an automated nurse-call button in a way. It tells the caregiver they need to go check on the patient. Video could provide more data, and I'm not saying video clips would not be helpful, but for this system it is less about using video for remote monitoring, and more about using video analytics as a way to better analyze the scene and determine when to ring for a nurse/caregiver.

Wouldn't that get old quickly with a false rate so high?

Undisclosed #1

IPVMU Certified | In reply to Brian Karas | 12/07/16 05:08pm

Sure you have. Burglar alarms, people counters using beams or motion detectors, and so forth. Lots of event systems do not have video data to go with the event, and really only tell you a sensor triggered, but do not give any details of how/why the trigger happened.

Fair enough.

But, IMHO, if you had a burglar alarm that had 25% false alerts, you would be real interested in the how/why.

And sure, lots of 'event systems do not have video data to go with the event', but do lots of video event systems not have video to go with the event?

Even so the system may be useful; it just seems it would be a lot more useful if a standard replay of the event was provided. Since they are apparently unaware/unconcerned of HIPAA implications, I'm wondering why they wouldn't offer it.

Jack Sink

IPVMU Certified | 12/07/16 03:38pm

The key resistance by providers I've met with regarding in-room video is the storage of video can allow baseless claims of negligence by families looking for a reason to sue, and equally, knowing that stored video could give basis for a claim due to misbehavior or abuse of a resident by caregivers that can be discovered through review of stored video.

The limitation of live video only might reduce that concern by operators but still yield a desireable tool for respondimg to possible problems.

Undisclosed #1

IPVMU Certified | In reply to Jack Sink | 12/07/16 03:44pm

Maybe a twenty minute retention?

Scott Sheldrake

12/07/16 03:40pm

They should make an analytic that detects when companies like this run out of investor money and fail. Data could be send to the unemployment office for expediting applicants or used by police for early warnings of new MLM startups.

Read this IPVM report for free.

This article is part of IPVM's 6,807 reports, 913 tests and is only available to members. To get a one-time preview of our work, enter your work email to access the full article.

Already a member? Login here | Join now
Loading Related Reports