Avigilon Video Analytics Product Manager Quits

With the VideoIQ acquisition complete, Avigilon has made a significant hire to product manage their video analytics portfolio. Instead of promoting / moving someone from VideoIQ, they have a hired away a SAP veteran who managed business intelligence.

Here's a short demo of SAP's business intelligence approach:

Notice this is similar to retail video analytics but much different than security ones.

It will be interesting to see how this evolves as 'data mining' and 'number crunching' have been far less popular solutions for video analytics in surveillance that tripwire / perimeter violation and other alert focused actions.

Update: the manager has now quit.


Man, it seems business analytics and video analytics have very little crossover. In the intro video embedded above, there is no mention, not even a hint, of video.

I suppose at some point data is data, but doing video analytics right and understanding the challenges behind it are really specialized.

This may be a great hire, but plumbers don't typically make good electricians or vice versa.

I'll play devils advocate and make the case for it:

Video is just video, until you can turn it into actionable data that improves the business. It's only when you do that, that you can maximize its value and profit making potential (for both supplier and customer).

This person knows how to do that. VideoIQ people do not.

Now, of course, the question becomes how will they turn video analytics into business intelligence. In what applications? How well will it work, etc.

I welcome this marriage. VideoIQ was expert as using VA to improve security. Maybe a fresh perspective can help identify where else VA can provide value. With any luck, this team may hit on a host of new applications that have tangible ROIs for VA, unlike security. Good luck to them!

SAP being a leader in enterprise data management, I can definitely see how this can be a plus, particularly if Avigilon develops an application that will allow customers to interface the data from the VideoIQ appliance with SAP and other ERP tools so it can be manipulated and actioned easily.

They'll be able to bring immediate value to an A-list of well established, and usually large clients, as well as allow them to generate very interesting cross-selling opportunities for their other products.

What would be an example of a practical application where you would 'interface the data from the VideoIQ appliance with SAP'? I ask because I am not aware of any other analytics vendors doing such integrations. If you know of specific implementations, please share.

This isn't a video analytics example, but it provides a general idea for what could be implemented, minus the need to add dedicated people counting appliances since your video surveillance infrastructure could now be leveraged to accomplish that function.

I'm not a retail expert, but I'm also guessing that by using the video surveillance infrastructure to support this type of data gathering, you could get even more granular, say by department, sections in departments, or aisle in a retail space, since you would now be able to actually count and track the people going into and out of each of those areas at any given time of the day and tie that into staffing, POS, and other operational ERP modules, information related to promotions, etc...

The ERP system would be programmed and used to massage that data in different ways so that it could be turned into actionable information for management.

Many examples are provided about the extent to which retailers are going to leverage in-store traffic data in this recent article, including one where an application called Prism Skylabs is used to leverage security video data, so there's definitely some very good potential in that space.

But I digress.

What would be an example of a practical application where you would 'interface the data from the VideoIQ appliance with SAP'?

To answer your original question, SAP commonly interfaces with a number of data gathering devices used in different types of businesses. An API for the VideoIQ appliance should enable the two systems to communicate and data to be transfered so that it can be used by SAP or another ERP.

Alain, thank you for explaining to me what an API is. Again, my question is a practical application.

Your practical application is retail. The problem is there are lots of companies already doing this in retail e.g., Brickstream, Irisys, Shoppertrak. My goodness, there is one company in Vancouver, literally down the block from Avigilon, that does exactly this - Lighthaus.

So Avigilon bought VideoIQ, the most trusted security video analytic and is going to turn it into retail solution, when there are lots of optimized retail analytic solutions already out there?

The person they hired away from SAP was recruited to product manage, i.e. make sure the lifecycle of applications that already exist is managed properly and that other opportunities to leverage the technology are thoroughly investigated, including what the competition might be doing, and that the most promising ones get developed and make it into the portfolio.

Any data attributes related to video information - not necessarily the footage itself - like time, space, objects, movement, speed, direction, thermal carateristics, color, shape and textures can potentially be captured and manipulated in order to provide information about events and assist in managing them.

Retail is one of the most obvious applications, which is why I guess I just sort of jumped on it - aided by the fact that you mentionned in your original post - but there are many others I can think of off the top of my head.

Road and other types of traffic management, crowd assessment, manufacturing process control, materials movement, access control come to mind, but basically anything that would typically involve something you can see in action and need to assess can also be a candidate.

And you're welcome for the API...and the product management explanation, since they weren't only for your benefit. ;)

Avigilon is a growing business and you would expect they would take whatever measures are needed to make sure that growth is maximized and managed properly and I'd be disappointed if they didn't. Hiring somoene from the outside should help take that load off other VideoIQ personnel whose strengths are in actual product and application development.

"Road and other types of traffic management, crowd assessment, manufacturing process control, materials movement, access control come to mind, but basically anything that would typically involve something you can see in action and need to assess can also be a candidate."

So now you are just throwing things out. The challenge here is actually implementing these things. How is Avigilon (or any surveillance manufacturer) going to do these things? It's not like video analytics and associated applications are brand new. People have been working on this for years, with very little success in these areas.

"Hiring somoene from the outside should help take that load off other VideoIQ personnel whose strengths are in actual product and application development."

VideoIQ people actually have experience in application development. That's why they are not randomly suggesting things like crowd assessment, access control, etc.

It will be interesting to see. They acquire a company and then a month later hire an outside product manager to run it. I am not sure if VideoIQ personnel will see this as taking a 'load off'....

"The challenge here is actually implementing these things."

I don't mean to beat a dead horse, so this will be my final contribution on this line of thought, but the logic of having someone with solid product and project management experience will help make sure the people who are the best at doing this - the folks acquired with VideoIQ - don't have to worry about those details and can actually focus on development and implementation.

So the people at VideoIQ don't have solid product and project management experience?

At 3VR, I saw waves of really smart, experienced Silicon Valley outsiders come in to manage product / projects / marketing, etc. The problem repeatedly is that it is very hard to manage something that you do not understand the development and implementations details of because you lack sense about what can be done and what customers in the space want to be done.

"So the people at VideoIQ don't have solid product and project management experience?"

I didn't say that, but maybe they would prefer doing something else, which is usually the case for most scientific or technical experts from my past experience.

This comment reads like something Dan Dunkel would write.

Alain, no offense intended, but it does not appear you understand what SAP (and like systems) do in the retail world.

The best example is auto replenishment and analysis. A store sells 100 of X widget in a day. The store only sells 100, on average, in a week. The SAP system automatically adjusts that stores next delivery from the DC while also adjusting the order from the manufacturer. It also flags this to the buyers and compare X widget sales across sister stores to identify any type of trend they may be able to get a jump on. This information is also shared with merchandising, operations and marketing.

Given this scenario, how do you think video analytics would add any actionable intelligence?

What you're refering to are agregate sales, inventory, and order management. Those are only a few of the many things SAP and other ERP systems can do for a retail business.

With traffic analysis, which SAP can also handle through the BI module, you can actually get down to how many customers walked into the store and when, how many of those walked by the display where those widgets are and when, how much time did those customers spend in front or in the area of that display, when was each unit sold, was there a promotion on when it happened, was that promotion for those widgets or something else that may or may not be a considered complimentary product, how many of those widgets were actually in the display when those walk-bys happened (were they out of stock or not), was there enough/too much/not enough staff to serve them at the time the customers walked by the shelf where the widgets are located so that the sales could/couldn't be closed, and so on.

Assuming all of that data has been captured using the proper tools, video analytics potentially being one of them, this is all extremely useful information for a retailer to have in order to manage its operations when you're able to put it all together so that it makes sense.

With traffic analysis, which SAP can also handle through the BI module, you can actually get down to how many customers walked into the store and when, how many of those walked by the display where those widgets are and when, how much time did those customers spend in front or in the area of that display, when was each unit sold, was there a promotion on when it happened, was that promotion for those widgets or something else that may or may not be a considered complimentary product, how many of those widgets were actually in the display when those walk-bys happened (were they out of stock or not), was there enough/too much/not enough staff to serve them at the time the customers walked by the shelf where the widgets are located so that the sales could/couldn't be closed, and so on

Customer counting / walk bys is the only item here that video analytics can do that cant be gleaned from the POS data. The store already knows how many they have on hand, and how many they sell and if there's a sale. What the widget is determines whether the customer needs assisitance, whether its answering a question or unlocking a case.

But as mentioned none of this is applicable to the SAP model. If it was, the Germans would have brought it into the fold long ago, as they did with RFID for inventory logisitcs

But as mentioned none of this is applicable to the SAP model.

I think SAP is very aware of the challenges retailers face, and as I pointed out earlier their BI module can already be used to address this.

The single widget example oversimplifies those challenges. Most retailers have hundreds, if not thousands of products to keep track of, spread across dozens of aisles or departments, and hundreds of customers who come in every day and who's needs they need to try to meet. Because when they don't, those customers are very likely to go somewhere else and those are lost sales.

Online retailers have been on top of their customers' buying habits, such as when they buy, what they buy, and how much time they spend shopping on their sites, for a long time simply because they have access to a lot more detailed information about them.

Brick-and-mortar businesses are just beginning their journey.

Net/net, Avigilon bought a security focused video analytics company for $31 million, hired a product manager whose experience is in a totally different market that is unproven for security/surveillance vendors to succeed in, and Avigilon dealers are excited...

Their new hire's experience extends way beyond retail. In fact, I don't even think she has retail experience. That's just the example we've been focusing on on this thread since yesterday.

Having someone from the outside, with a different skillset, doing this will probably could have more benefit than might be apparent at first because she's likely to be asking questions insiders might never have considered.

"Having someone from the outside, with a different skillset, doing this will probably have more benefit than might be apparent at first because she's likely to be asking questions insiders might never have considered."

In related news:

"Phil Jackson was hired as the head coach of the Denver Broncos today. GM John Elway declared, "What we really need is an outsider who can ask questions us football people have never considered. Knowing the domain one works in is overrated, as Michael Jordon's successful transition to baseball proved to us all."

"Just hire some electricians to design and install your surveillance system. Their lack of experience is a big plus."

It seems pretty clear Avigilon wants to diversify its business beyond conventional security/surveillance and into other applications so that they can stand out from the pack.

And if I'm not mistaken, they've been saying this for a long time.

It seems pretty clear Avigilon wants to diversify its business beyond conventional security/surveillance and into other applications so that they can stand out from the pack.

Alain, you may want to sit down when you read this..... Many companies already package analytics like people counting, dwell and traffic flow into encoders, cameras, DVRs, HVRs, NVRs & VMS for retail already. This isn't to say their strategy won't work, but it certainly won't make them stand out from the pack

I do admire the loyalty Avigilon has inspired in you though.

"Alain, you may want to set down when you read this....."

I'll take that in stride. John also pointed that out in relation to retail in this thread yesterday.

"I do admire the loyalty Avigilon has inspired in you though."

I've been aware of them for a couple of years now and they've lived up to their advertising to date, <edit>although John has made it clear they overreach sometimes ;)<edit>, so I guess that has a lot to do with it.

That said, to John's and his team's credit, as well as many others like you who contribute to IPVM, I didn't know nearly as much about the security/surveillance industry a couple of months ago as I do today.

Evolution doesn't happen without risk taking. It's not as if they're throwing their security/surveillance expertise out the window.

"... as Michael Jordon's successful transition to baseball proved to us all."

But then you have guys like Deion Sanders, and a few other good multi-sport athletes. (Notice Micheal Jordan didn't make it into actually made that list ;) )

Update: Less than a year later, Avigilon's analytics product manager has quit Avigilon to join an HR 'datafication' startup.

Update: Avigilon has a replacement for the analytics product manager, Steven Mueller, who previously worked in computer networking.