What New Technology Should Integrators Sell To Drive Sales?

JH
John Honovich
Jan 21, 2017
IPVM

For many years, IP video technology was the industry's highest growth market segment. Integrators focusing on that had a great opportunity to grow. Now, things has changed as the race to the bottom has intensified and spread.

So what technology or product segment should integrators turn to now?

Here is a list, just to start the discussion, in alphabetical order:

  • Cloud / hosted video / hosted access
  • Deep learning analytics
  • Dones / anti-drone technology
  • Ground based radar
  • Robots

Any of these do you think are promising for generating real sales? Any other categories or tech you would recommend?

Your input would also be helpful in us determining where we focus our upcoming tests.

UM
Undisclosed Manufacturer #1
Jan 21, 2017

We have been working on a shift towards analytics and automation.

U
Undisclosed #2
Jan 21, 2017
IPVMU Certified

Smart home integration

MM
Michael Miller
Jan 21, 2017

I think the ones that are selling cameras are hurting but those of us selling solutions that solve customers problems are doing very well with business growing every year.  Analytics, Automation and Integrated Systems will see immediate growth while drones and robots are going to take awhile to catch on for various reasons. 

(4)
MF
Michael Fusco
Jan 21, 2017

Exactly. Selling solutions that solve problems and selling widgets are two entirely different things. The technology is a means to an end not the end ls itself.

so many people dont seem to get that and position themselves as "dealers" instead of "problem solvers".

what problems are our customers coming to us with that we are not able to solve for them using existing abailable tech - and that will tell us where we need to look..

(5)
UM
Undisclosed Manufacturer #3
Jan 21, 2017

Deep Learning (Machine Learning) will likely upend and disrupt nearly every aspect of IVA, and as such make very large impact on hardware design and VMS and PSIM UX.  

The added value some of the new analytics coming out can bring to even a small installation will escalate quickly this year so my advice as a VMS manufacturer to integrators would be to monitor which software companies are embracing early aspects of Deep Learning in ways that appear designed to increase the value of the overall solution you want to sell your customers.  

(1)
(1)
JH
John Honovich
Jan 21, 2017
IPVM

monitor which software companies are embracing early aspects of Deep Learning

Which companies are those specifically?

The problem I have with your proposal is that its so vague it is not clear how any integrator could generate sales in 2017 based on it.

(1)
U
Undisclosed #5
Jan 22, 2017

"Avigilon appearance search" I think using Deep Learning

U
Undisclosed #2
Jan 22, 2017
IPVMU Certified

Why do you think that?

JH
John Honovich
Jan 22, 2017
IPVM

"Avigilon appearance search" I think using Deep Learning

I believe you are right. Avigilon has not marketed appearance search as 'deep learning' yet but given how hot the term is getting, it would probably make sense to do so.

U
Undisclosed #2
Jan 22, 2017
IPVMU Certified

Why do you believe it is using 'Deep Learning'?

Is it something about the way it works or something about the way it is described or something someone has said about it?

U
Undisclosed #5
Jan 22, 2017

All of the above :)

(1)
JH
John Honovich
Jan 22, 2017
IPVM

I can be more specific than #U5.

Here is a description from an Avigilon engineering intern:

Also, Avigilon has confirmed that existing Avigilon HD NVR Premiums will require an add-on kit that "contains the NVIDIA Quadro M2000 GPU".

(1)
U
Undisclosed #2
Jan 22, 2017
IPVMU Certified

An important distinction needs to be made between

1) a company offerring products that actually do deep learning 

2) a company who performs deep learning on their own servers and then use the rulesets that are statically generated in their products.

The difference is important as the models used in the first scenario must be somewhat generic, while only in the second scenario can they be truly adapative to the customers use case.

Based on the interns cv, it sounds as though it is currently the first case.  

Perhaps systems with this M2000 add-on kit will actually "learn", but appearance search as it stands today may just be well "taught".  IMHO.

UM
Undisclosed Manufacturer #3
Jan 23, 2017

As many of the below commenters already starting pointing out good examples of what I meant by keeping an early eye on who'd doing what with deep learning and how it will help integrators add significant value to the overall surveillance solution they're offering.  

There will be others this year introducing new features to software driven by DL.  NVIDIA's massive booth at ISC I'm sure will showcase some amazing facial recognition and inference based video analytics.  Maybe some autonomous security drones or computer vision/resolution advances etc.  

Essentially that's all I meant by keeping an eye on whoever is promoting new deep learning features.  In the beginning most will be designed to give real feature differentiation which will likely benefit early reseller adopters by providing them some really demonstrable value differentiation.  And one thing agreed on by pretty much every tech company in the world right now is that deep learning works, and works very well when using it to change old norms.

(1)
(1)
JH
John Honovich
Jan 23, 2017
IPVM

And one thing agreed on by pretty much every tech company in the world right now is that deep learning works, and works very well when using it to change old norms.

They agree it works in their application. They neither have the interest nor the experience to take a position on whether it will work in / for video surveillance.

I am not saying DL will or won't, I just don't general tech interest in DL is a strong reason to accept DL. Let's see what comes to bear in video surveillance, how well it works, how much it costs, what complexity it takes to add and than we can assess its impact.

U
Undisclosed #2
Jan 21, 2017
IPVMU Certified

Deep Learning is not yet Deep Earning

(1)
(4)
UI
Undisclosed Integrator #4
Jan 22, 2017

Buzz word selling

JH
John Honovich
Jan 22, 2017
IPVM

Buzz word selling

#4, can you be less cryptic? Are there any new(er) technologies that you are interested in?

UI
Undisclosed Integrator #4
Jan 22, 2017

John,

As a technologist I follow all technical developments with great interest.

But as a businessman, I am only interested in employing the tech that helps me solve my customers' problems. 

And what do those problems tend to be? well, for the most part, they aren't all that exotic - they want clearer ( especially at night), more usable footage for longer periods of time.

"Machine Learning", "analytics", "cloud", etc all sound great to industry insiders but clients could care less. Leading with that stuff is futile.

I get more excited about the development of things such as h.265, improved SSD storage, improved camera sensors and light grabbing tech - these things directly impact core client demands to view their video at the highest frame rate, resolution and image usability that current tech allows. All else is secondary..

(2)
(1)
(1)
MM
Michael Miller
Jan 22, 2017

What good is the best image quality of it takes forever to search for what you are looking for?  "Machine Learning" and "analytics" allow you to search vasts amount of video in seconds.  In my experience customers that use their system every day are very interested in this technology. 

(2)
(1)
UI
Undisclosed Integrator #4
Jan 22, 2017

Sorry, Image quality will always trump analytics. What good is an image that can be brought up in two seconds if the license plate number cant be read, the face is a blur or the frame rate caused a gap at the moment of truth? the core purpose of the system always takes precedence. Analytics and such will always be secondary. 

Dont take my word for it - go out and try to sell a customer a system with crappy low light imaging recording @ 8FPS but with wonderful machine learning and analytics and let ,me know how it turns out!

(2)
Avatar
Jon Dillabaugh
Jan 22, 2017
Pro Focus LLC

So you must only ever sell 30MP Avigilon Pro cams?

What good are crystal clear images if you have to go back and review to find actionable info? My client needs awareness that analytics hopefully can soon deliver reliably. It's hard to watch hundreds of camera streams looking for a needle in a haystack. But analytics can easily manage that task. 

UI
Undisclosed Integrator #4
Jan 22, 2017

I'm not saying analytics doesn't have a place, it does, primarily in the scenario you describe in larger systems - I'm saying that the industry shouldn't get ahead of itself and start leading with secondary feature-sets and industry buzz words instead of focusing on the constant improvement of the core purpose of the system, which is and always will be - the image. Image quality drives all other needs.

I have found that in systems I have sold with high-performance specs, the client gets an immediate and ongoing ROI by having actionable images that  in turn spurs them to look to further invest in the solution, and that's where the analytics come in. Nobody is going to invest in a system that isn't effectively delivering on its core purpose. Except government perhaps. 

(2)
MM
Michael Miller
Jan 22, 2017

I think the image quality gap has shrunk between major camera manufacturers and will continue to do so.  If you know how to configure the image quality and compression in a camera you have a number of camera options to pick from to meet your customer's needs.   It's what you can do with this video is what is exciting and the future. 

Avatar
Jon Dillabaugh
Jan 22, 2017
Pro Focus LLC

My biggest client is begging for reliable analytics. 

(1)
MM
Michael Miller
Jan 22, 2017

Jon,  What is your customer looking to do with analytics? 

Avatar
Jon Dillabaugh
Jan 23, 2017
Pro Focus LLC

Reliable perimeter protection is the first goal, but intelligent "Google" searching would be nice too.

For instance, a child wearing a blue shirt goes missing, to be able to search quickly for "child blue shirt" would be awesome. Would save tons of investigative time searching grounds instead of a few clicks of a keyboard.

Note that when I talk about these features, I want them to work very reliably. With out the high reliability, they won't be utilized enough.

(1)
U
Undisclosed #5
Jan 23, 2017

Jon,

just for fun I Google "child blue shirt"

excellent results

(3)
JH
John Honovich
Jan 23, 2017
IPVM

just for fun I Google "child blue shirt"

excellent results

Please stop these types of comments. They do not help the discussion. I am not sure if you are trying to be funny but please try to be serious and explain your thoughts with at least a few sentences.

(1)
MM
Michael Miller
Jan 23, 2017

This is what Avigilon's Appearance Search is designed to do. Right now you have to find the person in the video to search for them across all your cameras but I could see them adding a search option say " show me people wearing red" or "show me green cars".  

CamIO can do this but from what I understand you have to have a WAN connection that will support all of your cameras that you want to include.  I think right now this is geared towards 4-8 camera systems, not 100+ camera systems unless you have a really big upload speed.

Briefcam can also do this but I don't think you can search across all cameras in the system like you can with Avigilon or CamIO. 

[NOTE: Further Camio discussion moved here - Using Camio With VMS / Larger Scale]

(1)
Avatar
Ricardo Souza
Jan 22, 2017
Motorola Solutions • IPVMU Certified

I agree. 

Reliable analytics with a high percent of precision and automation still has a long way to go. Deep learning analytics has been around for decades, it is certainly picking up the pace, but i would give it some more years;

(1)
JH
John Honovich
Jan 22, 2017
IPVM

Deep learning analytics has been around for decades

Ricardo, that is simply factual false. Any type of video analytics has only been around a little more than a decade and none of them were deep learning / neural networks until quite recently; and those deep learning efforts are still quite limited.

U
Undisclosed #2
Jan 22, 2017
IPVMU Certified

...and none of them were deep learning / neural networks until quite recently;

Except maybe for BRS Labs, who have been claiming to use neural nets for many years now.

JH
John Honovich
Jan 22, 2017
IPVM

That is true about BRS Labs though BRS Labs claimed many, many things and delivered on basically nothing, so it is hard to count them as an indicator of deep learning, one way or the other.

JH
John Honovich
Jan 22, 2017
IPVM

I get more excited about the development of things such as h.265, improved SSD storage, improved camera sensors and light grabbing tech

I agree that those things have use. Certainly to that end we cover smart codecs, H.265, camera sensors and improved low light more than 'buzzwords'.

However, in terms of breaking into markets or generating really high growth, this often comes from new technology / product categories that spurs users to increase their overall spending.

Avatar
Christopher Uiterwyk
Jan 22, 2017
IPConfigure

Secure/encrypted end-end security solutions will be the dominate topic and opportunity with organizations that have formal IT departments.

(3)
UI
Undisclosed Integrator #6
Jan 22, 2017

I don't think this is going to be a big growth driver. This was/is already pretty much addressed by selling an end to end like Avigilon / Honeywell / etc. I don't really think there is a lot of extra growth to come out of it.

I think the commentor above was right when he said that the high growth needs to come from new product categories. The reality is that video is going to be a low margin market for the foreseable future. Few people are really going to pay 4x a camera for analytics and the price difference between analytics and no analytics is going to become minor in the next two years as the asian manufacturers include it more often into their mid and low range camera lines.

I really think the growth with regards to video will come from video services (digital guard tours, video alarm, etc.) because those services can compliment or take the place of physical security (guard patrols etc.). I think the next problem the industry is going to run into is that access control is setting up to become a race to the bottom as well since the VMS manufacturers are now starting to pivot to that industry.  The redeeming thing in access control is that if you do the door hardware as well, that industry is not going to change significantly simply because there are new entrants pushing cheap door software and door controllers. However, if you haven't embraced IP locks or the retrofit solutions from companies like Salto, I would pivot now.

IP locks plus build out of the more mature access software for the larger companies into more compliance and safety management is going to drive growth at the top of the food chain. The problem is, the majority of the integrators that make a living in this industry are smaller fragmented companies that do not have the staffing ability or expertise to sell nor up sell Fortune 500's and similar enterprise customers spread across the US or Globe. So, the big boys like Stanley, Convergint, etc. will probably take an ever larger portion of the growing pie at the top end while mid and low end integrators will see anemic growth with the exception of those embracing video services.

(1)
(2)
Avatar
Brandon Knutson
Jan 23, 2017
IPVMU Certified

As an end-user, every place I have worked as a security manager has wanted to add "card readers" to numerous doors, but they do so very selectively because of cost. This market will remain slower than it should as long as a lockset costs $300 and a card access door costs $3,000. Figure out a way to add a door into an ACS for sub-$1,000 that isn't battery powered and this segment will grow. A good ACS is a great problem solver for facilities and security peeps.

My other recommendation to integrators... Regularly let your clients know what you can do for them; do not wait for them to call you. This sounds obvious but I have seen this happen to me three times in 25 years as a security manager.   

UI
Undisclosed Integrator #6
Jan 22, 2017

Being a dealer of rs2, s2, lenel, Avigilon, Dsx, Bosch, Dsc, dmp, and Honeywell I have to say I am most excited about Honeywell due to the xtralis and videofied acquisitions. I think the digital guard tours and video alarm monitoring are going to be strong add-on services to combat the loss of video margin. Excited for all the access control platforms to get updated gui's and also excited to expand our selling of IP lock sets and credential on card for interesting applications with salto padlocks and cabinet locks.

(2)
Avatar
Ricardo Souza
Jan 22, 2017
Motorola Solutions • IPVMU Certified

Deep integrated and Automated solutions are key. The future is calling. Operators are too much susceptible to work conditions, stress, attention span, etc. AI can work in surveillance much better than we.

I also have high hopes we will be seeing automated Drones with cameras flying over us very soon.

Just recently, US Military conducted successful tests with drone swarms. And those just need a bunch of programming, a set of instruction and a GPS signal.

Avatar
Timothy Howell
Jan 23, 2017
IPVMU Certified

As an engineer with a graduate degree in psychology, I am interested in the deep learning analytics issue.  Frankly, I think it is way too early to begin selling in this area.  As I attend conferences across the country and survey the literature on the topic, no one seems to have concrete ideas about building or security analytics.  If others are interested in this topic I will post what I have learned after I finish up my investigation.  For now, I am not seeing how these unnamed and undescribed "analytics" are providing any contributions.  I will be glad if someone can show me where I am wrong here, I think long term it will be a good deal for integrators however manufacturers must develop tool sets that can be deployed with VMS's and provide training for integrators... I am not seeing much of that yet...  Thoughts?

(1)
(2)
UI
Undisclosed Integrator #7
Jan 23, 2017

Improved coverage with P2P radios. Easier MESH network design and more affordable. Cheaper ways to store video for longer periods.

(1)
Avatar
Sean Nelson
Jan 23, 2017
Nelly's Security

automation products/services are super hot right now, get in while the gettins good

JH
John Honovich
Jan 23, 2017
IPVM

Sean, how do you see automation sales trends vs video? Automation bigger? Growing faster?

Avatar
Sean Nelson
Jan 23, 2017
Nelly's Security

speaking of our business personally, other than 2GIG, we dont have a dedicated home automation hub. With that being said, we are still doing fairly decent with the zwave products we stock that works with alarm.com and 2GIG. But I am looking heavily for a dedicated home automation hub to stock that doesnt require monthly subscription service to get in on this bonanza. 

If I was an integrator though, I would definetely be offering this as a service. Seems people are fascinated with Alexa, Smart Thermostats, Locks etc. This used to be dominated by the high priced solutions such as Control4, Crestron, and such. Those systems could only reach a certain clientele. Now their are more budget friendly systems out their that are rapidly closing the gap in performance as compared to what you can pay for one of those high priced systems that doesnt cost a small fortune. IMO, this is an easy sell. People think its fun, and its easy to sell fun and convenience.

(1)
MM
Michael Miller
Jan 23, 2017

Sean check out Vera "No Monthly Fees.  No Contracts. No Hassles"

 

Avatar
Sean Nelson
Jan 23, 2017
Nelly's Security

I actually have one, havent played with it in a while, but have heard they arent very friendly to smaller distributors such as us. Like they want wal-mart type customers. Do you have any experience dealing with them? Just sent in a request to them.  

Avatar
Bob Kusche
Feb 20, 2017

This may be perceived as self-serving, but remember that you asked, John!

Based on our personal experience, SpotterRF ground-based radar has made a profound difference in IPVS-related sales here at NMS Security. We are a distributor for SpotterRF; we also offer an optional plugin for Milestone XProtect, allowing radar-generated movement trails to be viewed, recorded, and handled exactly like a camera feed. In fact, our plugin received an award at last week's Milestone MIPS conference.

Why is SpotterRF hot? Verticals with very large areas of land where human, drone, or animal detection is required do not have any other cost-effective tool to cover more than a perimeter. At a list price of $6k to $45k, each IP-networked solid state panel can cover from 1 to 260 acres and 30 simultaneous moving objects. The system will direct multiple PTZs to view the moving object(s). Sophisticated filtering can include or exclude objects based on unlimited custom-drawn zones with size, direction, speed, displacement, duration, total distance (and more) dictating whether an alert is sent.

Check out http://www.nerc.com/pa/Stand/Reliability%20Standards/CIP-014-1.pdf for a mandated directive for electrical substations to improve their physical security. Airports, water utilities, data centers, mines, bridges, and any drone detection requirements are other especially strong verticals.

Locating, installing, and supporting a radar solution is very similar to IPVS. A three-day course and three installations are the pre-reqs of certification. Typical sales range from $75k to $__M.

We encourage any integrator to go to www.spotterrf.com and contact me if they would like more information on how to get started.

(1)
GM
Gordon McFarland
Feb 20, 2017
Bowler Pons Solutions Consultants

We have been integrating automation with surveillance, access control and PSIM.

In addition, we are offering information security assessments of the network itself.

New discussion

Ask questions and get answers to your physical security questions from IPVM team members and fellow subscribers.

Newest discussions