Dissapointed? The level of expectation.

desillusieStrange,

everytime if I talk about Video Content Analysis most people try to convince me by the fact that it wont’t work! VCA will result in a desinvestment. It will not do what is prommised…

Ever heard about the level of expectation? When the CD was introduced by Philips & Sony, the people expectation level was that the content stored at a CD was of a high level and that it didn’t had any expire date. Burning backup’s on a CD was the future. Just a few years after the first introduction people where confrontated with the weakness of the CD. If there was a scratch on the disc, sometimes it didn’t work anymore. Worse, after a few years some disks had CD rot which made them useless… If the CD was filled with very worthfull pictures of your little child…. or if it was filled out with beautifull memories of your mam or dad you can imagine how you would feel about that! First the CD was the fututre but just a few experiences later you where aware of the fact that you had to handle it with care.

Within VCA we see the same level of expectation. When VCA was introduced, we thought that this was it! With VCA we could do everyting. We even could do much more with less people etc. But what about rain, snow, clouds, spiders and so on?

To make use of VCA we have to aware of the fact that it is a great solution if we make use of it in the right environment. In the automotive industry we see VCA in a car filled out with lane assist. The camera in the car screens the road in front of you and keeps you driving just between the lines. Also the speed control zones where your speed is messuered over a long distance is filled out with LPR (License plate recognition) systems so the ticked can (and will) be send over to you.

There are lot’s of usefull situations which we can fill out with VCA. The only thing we have to do today is managing the level of expectation in the right way. As we wrote down in our vision,… VCA will be the future.

Kind regards,

Norbert

Advertisements

8 Responses to “Dissapointed? The level of expectation.”

  1. Ruud Says:

    Hi Norbert,
    Great to see that there is a IVA, VCA or iVCA what’s in the name 😉 blog!
    I agree by reading that the level of expectation needs to be managed so we don’t have to sell the solution twice. Once the level of expectation and once the product itself.

    Keep up doing the good work!
    Cheers, Ruud

  2. Greg Cortina Says:

    Norbert,

    I could not agree more. As the industry tries to tame the original beast the users are asking for more. More behaviors, more reliability, more camera input types, panning camera analytics. Where will it end?

    Vidient tackled a very intensive customer problem recently providing “person in tunnel” detection using existing cameras with less than perfect video quality.

    Managing the customer expectation was certainly a focus as we have seen many crash and burn upon deployment. However, with the assistance and patience of the customer we created a behavior that reliably detected intrusions while a train was moving in the background and a person walked along side.

    At the same time we have been asked to do the impossible (or at least extremely unlikely) and have happily passed on project. The customer expectation was too high.

    What really brings us to a successful conclusion is the direct communication and assistance the analytics creator maintains with the end users. We could not have accomplished that task with an “out of the box” solution and fortunately our business plan is to have our engineers available.

    Your article is timely and accurate.

    Greg

  3. Viktor Ollari Says:

    Maybe the main problem is the wrong marketing. 🙂
    Most of the Video Analytics developers project to the market: “you can do anything with our solutions”(instead of “you can use it here, and there, with the following results:…”)

    The result of projection marketing: high level of expectation —> disillusion, disappointment

    • Norbert Says:

      @Viktor Ollari,
      I do not agree by that. Within my job as application manager my focus area is based on Video Content Analysis. Today the main manufacturers don’t visably positioning their product at a level which is above reality. My experience is that mostly the customers level of expectation is based on the ‘stories’ out of the market and not on the marketing philosify as presented by the manufacturers.
      What we have to do is to get rid of these stories not by telling that these stories are not true, but by focussing on the real VCA related possibilities. As I wrote in my article, license plate recognition for example in combination with speeding or Lane assist as used in the automotive are some good examples of VCA solutions which shows us proven technology.

      Kind regards,
      Norbert

      • Viktor Ollari Says:

        Norbert,

        probably, I put my words in a wrong way, but on a certain level we agree. I mean you wrote about Lane assist, LPR (“by focussing on the real VCA related possibilities”), while I wrote: where and what (kind of solution) to use to achieve certain results.

        As I see there are three main seed of disappointment.

        First problem: You’re right, there are ‘stories’, so called MI (Mission Impossible) stories on the market and expectations based on them. And it’s hard to tell the customer “the mission you want is impossible”.

        Second problem is: At (or from?) the begining of “VCA era” the several (even well known ) developers said (only for example): “our product is self-learning”. You are right: it’s not above reality, just part of the truth. Because, self-learning is not equal to human type AI, but in most of the customer’s mind it is. Misunderstanding, that is usually not handled by manufacturers (there are some exceptions), because it’s useful for the better marketing results. Magic of words and understanding (marketing).

        Third problem: false alarms. Abandoned Object, for instance. I tested some solutions (built is and softwares as well) . They worked pretty well. But even the best generated false alarms (sharp edged shadows, strong sunlight patches, etc.). The marketing materials (ads, flyers) do not mention it, of course (that’s not their job), but a problem that must have been handled.

        So: MI stories, misunderstanding, false alarms. The result is dissapointment.

        You’re right! The way to solve the problem is focusing. Plus: facing the problems as well.

      • ivcablog Says:

        @ Viktor,

        quote (and comment):
        First problem: You’re right, there are ’stories’, so called MI (Mission Impossible) stories on the market and expectations based on them. And it’s hard to tell the customer “the mission you want is impossible”.

        comment: Why is it hard to tell the customer that the mission is impossible? Isn’t that exactly what they expect from us? Isn’t this excactly where things are going wrong?
        Just like i said before…. manage the customers expectation! That is what we stand for 😉

        Second problem is: At (or from?) the begining of “VCA era” the several (even well known ) developers said (only for example): “our product is self-learning”. You are right: it’s not above reality, just part of the truth. Because, self-learning is not equal to human type AI, but in most of the customer’s mind it is. Misunderstanding, that is usually not handled by manufacturers (there are some exceptions), because it’s useful for the better marketing results. Magic of words and understanding (marketing).

        Comment: I ‘agree’ by that.

        Third problem: false alarms. Abandoned Object, for instance. I tested some solutions (built is and softwares as well) . They worked pretty well. But even the best generated false alarms (sharp edged shadows, strong sunlight patches, etc.). The marketing materials (ads, flyers) do not mention it, of course (that’s not their job), but a problem that must have been handled.

        Comment: What about the discussion of a so called false alarm. The whole security industry is based on taking care of for about 96% of ‘false’ alarms. Isn’t it true that having an alarm to much is less important then missing just that one real alarm? Where did the mindset start where people starts to think that a new solution will have a fals alarm rate of 0?

  4. Bart Verhulst Says:

    As was mentioned regularly on recent the European VCA Conference from IMS Research in Brussel managing customer expectations is an important topic.
    There are a lot of possibilities with and for VCA solutions in different applications. But there also huge differences in performance, quality and price for different solutions and still you will find too many parties on the market that stress mainly the nice possibilities, shown in almost ideal circumstances and thus confusing the customer. Deploying video analytics e.g. for perimeter detection assumes competences in VCA, cctv, thermal camera’s,…. The scene changes constantly and is no laboratory environment mostly shown in marketing material. You must take care with a lot of factors among them FAR as well as missed positives.
    We would welcome more VCA standards and quality labels to be spread (e.g. i-LIDS) and more training to be given to end users and consumers about what is possible, and what isn’t in order to set expectations more realistically but also to make it easier to compare and understand differences. We also pleed for manufacturers to be open and complete in marketing information. Detecting left objects in an empty train station is not really how a real live train station is looking, is it? Detecting an intruder on a fence line with a camera view from outside the fence line is not realistic either, is it? If you have VCA available at the edge as on option on an a standard IP camera or encoder and you do not mention that the VCA analysis can take up to 60% of the streaming performance, your creating somehow any confusion isn’t it?
    Yes we can deliver good VCA solutions for various applications, but circumstances are not alike, just as performance, price of VCA systems as well as compentences of the integrator will vary.

  5. Viktor Ollari Says:

    Norbert,

    “Why is it hard to tell the customer that the mission is impossible? Isn’t that exactly what they expect from us?”

    It’s always hard to demolish a dream… 🙂
    Even in buying so objective product like security there are feelings and dreams above realism. There are exceptions, naturally.

    Well, I think they expect the fulfilment of their needs (desires).

    “manage the customers expectation! That is what we stand for”

    I agree. 🙂

    “What about the discussion of a so called false alarm.
    …”

    The problem is the inconsistency between marketing messages and the real world. Yes there are a certain level of false alarm. It can’t be avoided.

    But, several VCA developers say: ‘You need less security personal’. It’s true within a certain (project specific) level of false alarm, above it the same number of SP is needed. It’s usually not told to the customers.

    “Where did the mindset start where people starts to think that a new solution will have a fals alarm rate of 0?”

    Where?
    1.) In the ‘Accountancy rooms’, where the (security) budgets were decreased. The easiest way to solve this problem : kick off SPs. In this case ‘You need less security personal’ sounds extremely good.
    2.) In ‘Marketing rooms’. As I described above.
    3.) Faith. Security managers are usually not IT or video specialists. They believe in marketing leaflets, sales persons statements, MI myths, etc.

    So, I agree: ‘manage the customers expectation’! But VCA must be demystified, as a first step, and educate customers in the same time.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: