YT YouTube2Text

YouTube2Text → Transcripts → The Connected aquaManager Ecosystem: The next generation of production control - AI Launch Event

The Connected aquaManager Ecosystem: The next generation of production control - AI Launch Event — Transcript

by aquaManager · 3,795 words · 651 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:02makes this even more powerful and
0:05more exciting.
0:06Um it is something that we have been
0:09working on for almost 2 years and we are
0:12very happy and excited to present it to
0:16you today. So, let's go to the next
0:18slide. Don't worry, the presentation is
0:21a very brief one. It's only five slides
0:23and then we go to the
0:25to the system.
0:26So,
0:28before we show you uh what is new,
0:31we want to make one thing very clear.
0:34Everything you already have in
0:36AquaManager remains the foundation of
0:39what you are going to see today.
0:42I mean, your production structure, your
0:45historical data, your reports, your
0:48planning, your daily records,
0:51all of this is still there. Nothing is
0:55lost and nothing is replaced.
0:57It's only improved and what changes is
1:01what you can now do with it.
1:05Um
1:06really, what makes this possible is the
1:09strong AquaManager production backbone
1:12and the data model. This allows
1:16production information, real-time data,
1:18BI and AI to work together into one
1:22system in a meaningful way. So, this is
1:25not a um
1:26a new system starting from zero. It is
1:29the AquaManager you already know, but
1:32with more power, more connection, and
1:36more ways to help you
1:38uh uh
1:39control and improve uh production.
1:42So,
1:44let's see what is new.
1:48If I had to say it with one phrase,
1:51everything works together now.
1:55And what I mean,
1:57AquaManager remains the system of record
2:00for production management.
2:02Then business intelligence
2:05gives every level of your company
2:08the reporting they need
2:10from farm
2:12teams to top management.
2:15It is the management information system
2:18your company has always needed.
2:21Then we go to Aqua 4.
2:23And Aqua 4 gives you the live picture
2:27from the farm. So
2:30real-time signals and video streams are
2:33connected to the cages, to the fish
2:35groups, and to the production reality.
2:38And finally
2:40the AI assistant
2:42allows you to ask questions
2:46investigate faster and find
2:49what needs attention.
2:52This means you don't have to jump
2:55between different tools or search in
2:59different places to find the information
3:02and find out what's happening.
3:04Production data, real-time data,
3:07business intelligence, AI are now part
3:11of the same ecosystem. And the result is
3:14simple.
3:15One connected view of your operations
3:18from daily production
3:20to management
3:22decisions.
3:24Uh
3:25let's take a very quick look at Aqua 4.
3:29This is
3:31where you can see
3:33the live farm inside Aqua manager. Aqua
3:364 is the real-time part of the of the
3:39solution.
3:42Production reports tell you what
3:43happened. Aqua 4
3:45helps you, allows you to see what is
3:48happening now.
3:49So you can see oxygen temperature,
3:52equipment status, camera streams, uh
3:55fish behavior, pellet detection, feeding
3:57activity, all of them in one system.
3:59But,
4:00the important point is not only to see
4:04live data.
4:05The important point is that this live
4:08information is now connected with your
4:11production data.
4:13A sensor value is not just a number.
4:15A camera image is not just an image.
4:18It belongs to a cage. It belongs to a
4:20fish group, to to a feeding plan, to a
4:23production history.
4:25So, the system can work
4:28with both
4:30IoT data and actual production context.
4:34Not just oxygen is low, but oxygen is
4:37low for this cage with this fish density
4:41and this size of fish.
4:43And this is exactly where the real-time
4:46information becomes more useful.
4:51Because with Aqua Forum, you can see
4:53what is happening now,
4:56understand which fish are affected, and
5:00know what really needs your attention
5:03and your quick response. This is not
5:06just monitoring. It is real-time farming
5:10information connected to production
5:13control.
5:15And um
5:16now, let's talk a little bit about AI.
5:21As you know, Aqua Manager
5:24contains a very
5:25rich
5:26and very detailed production data model.
5:30The AI helps you unlock the power of
5:34this data. So, instead of asking,
5:38"Which report should I open?"
5:40or where where where can I find this
5:42information? You can start with us by
5:45asking questions. For example,
5:47which cages need attention today?
5:51Which cages had the highest mortality
5:55last month?
5:56Or
5:57where is feeding deviating from the
6:01approved quantity?
6:02How does growth compare across sites or
6:05across unit groups or across batches?
6:08Or you can ask questions related to the
6:11status and the health of the equipment
6:13like
6:15which sensors have not sent any data
6:18during the last hour?
6:20Or you can ask more advanced questions
6:22like which cages had dissolved oxygen
6:26below 5 mg per liter for more than 4
6:31hours per day during last week?
6:34The important point is that
6:36the assistant does not only
6:39make the retrieval of this this
6:41information possible and easy.
6:43It can help
6:45you to combine production data, feeding
6:48data, mortalities, growth, biomass,
6:50everything into one place so you can ask
6:53more advanced questions and get answers
6:56that
6:57would normally be too difficult, too
6:59time con- too time consuming, or
7:02sometimes impossible to produce.
7:05So, this
7:07helps you move faster from just data to
7:11knowledge, from knowledge to
7:13investigation, and from investigation to
7:16action.
7:18In very few words,
7:20your people bring the experience, the
7:22assistant
7:23provides the help they need to use the
7:26full value of the data that you already
7:30have. So, enough talking. I think here
7:33we can just go to the demo.
7:37Uh we don't have much time anyway, so we
7:39need to be to be fast.
7:42And
7:43please uh uh
7:44share
7:46the new Aqua Manager environment.
7:53Here it is.
7:56Okay, thank you.
7:58So, this is the new Aqua Manager
8:01environment.
8:02It is still the Aqua Manager you know,
8:05don't worry, but with a fresh
8:08modern nice user interface and this is
8:12something that many of you have been
8:14asking for and
8:17of course we heard you.
8:19Before we get into the details,
8:22let us quickly show you only two
8:24examples so you can see the new look and
8:28feel. We will show you only the home
8:29screen and the custom dashboards. And
8:31what what you actually see is the home
8:33screen.
8:35It is a page that gives a quick view of
8:38the production information that we think
8:41is the most relevant or matters most for
8:44the majority of of the companies.
8:47As another example, let us show you the
8:50customized dashboards that allow each
8:53company to bring in one page, just one
8:56page, all the information that
9:00they think is the most important for
9:02them like feeding mortalities, FCRs, uh
9:05growth rates, cost, whatever. You can
9:08combine the information that you
9:09consider to be important is not in in
9:12just one page. And of course this is per
9:14user, so the the the CEO or the
9:17production director can have a different
9:19dashboard.
9:20Um
9:21from the
9:22production uh farm manager. Okay.
9:26Now, Neil, could you please go to the
9:28back to the to back to the home screen
9:31and let us take a look at feeding.
9:34What we see here? We see a difference
9:36between actual and approved feeding.
9:40Okay, nice. This is very useful
9:42information to know.
9:43But really, we want to go deeper. We
9:46want to understand why this happened.
9:49And
9:50we know again that many of you
9:54wanted
9:56faster answers without having to search
10:00for the information through different
10:03reports or screens or ask someone a
10:06person to prepare an analysis. And this
10:09is exactly where the AI assistant
10:13can help. So, let's go to the AI
10:15assistant.
10:18First of all, as you see, I don't need
10:20to log in again. I just move to the
10:22assistant and type my question. And the
10:25question in this case will be
10:27show me the feed deviations
10:30across April 25.
10:33Group the results by site and unit unit
10:35group and compare them with temperature
10:38and fish density. Just natural language
10:41you type to the system what you want to
10:44to get.
10:46Then the assistant is thinking.
10:49Thinking a bit more. Maybe it is tired.
10:54And at the end retrieves the data and
10:58prepares the analysis for me.
11:01Now, the important point here is not
11:04only that I asked a question in natural
11:08language.
11:09The important point is that the
11:11assistant can combine different types of
11:14information like feeding data, sites,
11:18unit groups, temperatures, fish
11:19densities
11:20that
11:22this would normally take time to
11:23prepare. Here, I can start the
11:26investigation quickly. I can see the
11:28results as a table, as a pivot, and as a
11:31sound. So, I can move quickly
11:35from a simple question
11:37to a useful
11:39analysis.
11:41>> Now, you know, Kostas, that very
11:43interestingly
11:44what we can also do with this engine if
11:46we want to even
11:49investigate a little bit deeper. So, I
11:50can ask a follow-up question. What do
11:53you think that caused this
11:56feeding deviation in the unit group or
11:59the site that we have identified? So,
12:01the system goes to the database and
12:03analyze analyze everything all together.
12:07Analyzes the density, the feeding,
12:10events that happened in the days or
12:13weeks before, analyzes temperature
12:16depletion, oxygen depletion, maybe
12:18temperature rise, and brings
12:20up to the table an an answer that makes
12:23a lot of sense to us as a growers and
12:27directs us to where we think that we
12:30where the system system thinks that it
12:32is clever to investigate. Now, when you
12:35have 20 cages, it is easier to identify
12:39yourself. When you have
12:41big number of cages, this thing helps
12:44you a lot to focus, to understand where
12:47you are.
12:48Back to you, Kostas.
12:49>> Oh, thank you very much, Nedu. That was
12:51a very very good point. And
12:54let me take the opportunity to add that
12:57uh
12:58these the results we see here are not
13:01produced by a generic chatbot.
13:04Uh
13:06we collected structured and injected
13:10aquaculture domain knowledge into the AI
13:14assistant. So, it is able to provide
13:18this type of of analysis.
13:20Uh again, thank you very much for your
13:22for your comment.
13:23Can you please go back to AquaManager
13:26and this time focus
13:29uh at another period, let's say January
13:312026.
13:36So, what we see here? What we see here
13:38is that there there were many
13:41mortalities during this period. So,
13:44again, the natural question is why? This
13:48time we will not go to the AI Assistant.
13:51We will go to Aqua 4
13:53because we want to investigate if this
13:56mortality was related to oxygen, O2
13:58environmental conditions, or, you know,
14:00other things that happened in the farm.
14:03So, we now move to Aqua 4. Again, no
14:06external login is required.
14:09And, as mentioned earlier, Aqua 4 is the
14:13real-time intelligent part of the
14:16system.
14:17Many many companies in the sector in in
14:20the aquaculture sector are not looking
14:23just for more sensors. They want live
14:27farm data connected with production
14:30reality. And this is exactly what we're
14:32trying to do with Aqua 4.
14:34So, it is a platform that I can see IoT
14:37data, equipment,
14:39cameras, everything. Let us show you a
14:41few examples. Nick, can you please go to
14:43Silver Fin?
14:45And let us show, for example, the this
14:47first page that is a sensor measuring
14:51oxygen temperature at the farm level.
14:54It's not placed in the specific uh cage.
14:57You see, I can see the results by day,
14:59by week, by month. I can define longer
15:03periods. I can set up the aggregation
15:06level. And it's actually very fast, even
15:09if I try to retrieve information for
15:11very long periods.
15:14In addition to the farm level, I can see
15:17the IoT data at the unit level. Imagine
15:20that I
15:21I put
15:22a sensor within each cage. Again, I get
15:26the same
15:27um access to information. I can see my
15:31feeding cameras as equipment.
15:34Can you please go back? Okay, feeding
15:36cameras as equipment. So, this is not
15:39the camera stream, it's the camera
15:40itself. And I can see diagnostics about
15:43the camera. So, I know, for example,
15:44what is the battery voltage, what is the
15:47um humidity within the camera control
15:49unit, things like that.
15:51I can see snapshots taken from uh from
15:54the cameras.
15:55Uh I can see the cameras live if
15:59I'm connected to them. I can see my
16:01stereoscopic cameras. These are the
16:03average weight
16:04estimation cameras. And um the results
16:08of the average weight measurements. All
16:10the information about equipment IoT is
16:12here.
16:13But, again, the real value is not just
16:18seeing live data.
16:20The value is connecting this data to the
16:25production.
16:26The cage, the batch, the the size of the
16:29fish density, etc.
16:31Which means we can ask more
16:35um right to the point questions. Not
16:37only oxygen was low, but was oxygen low
16:41for this cage
16:42and this fish size, and
16:44is this the reason for the
16:47high number of mortalities? So, Neil
16:49will try to show you an example on this.
16:52He's He's He will try to investigate
16:56the
16:57period of the period that we saw
17:00the high the spike on on on on
17:02mortalities.
17:04>> Now, what we identify here
17:06is that there is
17:09a spike of the mortality in that date.
17:13Okay. So, we will try to understand what
17:16might cause it.
17:18Um I will call the oxygen and the
17:20temperature. And I see something very
17:23interesting. I see the temperature
17:24depletion
17:25across the beginning of the month before
17:28the the the mortality spike. And if I
17:32will just zoom in a little bit here
17:35before the mortality,
17:37and
17:38I can see also oxygen depletion a little
17:41bit. Now, as fish farmers, we all know
17:43that this cause this might cause fish
17:46stress. This must might cause some
17:49problems with the fish. And this is
17:52well, almost a common knowledge. But
17:54imagine what could happen if you combine
17:57more information and you identify trends
18:00and you identify
18:02everything that happens at the farm also
18:05from the eyes of the measurements that
18:07you have on a real time at the farm.
18:10This is the superpower of this
18:13Aqua Four feature that we're presenting
18:15now.
18:17>> Plus the ability to to to to get to
18:20retrieve this data or or, you know, see
18:23what is happening using the AI assistant
18:26because there is an AI assistant also
18:28within Aqua Four. So, you can type
18:31questions like, "Okay, show me
18:32correlation between mortality and
18:34oxygen." Or show me
18:37mortality or fish appetite together with
18:40the number of hours that the cage was
18:42below 4 mg per liter of oxygen. Things
18:45like that. You can do amazing things. Um
18:48thank you, Nir. Can we please go back to
18:50Aqua Manager? And um
18:53we won't we will not show you a lot of
18:55stuff with Aqua Manager. I just want to
18:57say just a few words about data entry.
19:00And as many of you know, data entry can
19:02be done manually, can be done through
19:05integration with feeding systems or
19:07other applications, or through the
19:09famous mobile apps. And we only wanted
19:13to tell you that the mobile apps have
19:15also been redesigned and offer a lot of
19:18new capabilities and a new user
19:21interface. For example,
19:23with the new mobile apps, you can scan
19:24each individual bag of feeds. You can
19:27manage the full
19:29feed or consumables workflow after they
19:32receive the inventory. You can do a lot
19:34of stuff.
19:35In any case, once the data are in the
19:37system, AquaManager gives you many ways
19:40to use it. And the first
19:42option is to use the large set of
19:45already-made reports, like what Neil is
19:47showing here, which if I'm not mistaken
19:50is a mortality analysis uh by hatchery.
19:53So, you know, AquaManager provides about
19:55150 reports. You can use them and do
19:59almost everything you want to do.
20:03Now, another option to exploit the data
20:06is the BI layer, the business
20:08intelligence layer.
20:10And with the BI layer, you can define
20:13KPIs and metrics once, and then use the
20:17same trusted numbers across the company.
20:20Again, this is something that many
20:22companies wanted. One common reporting
20:26layer with trusted KPIs and numbers from
20:30farm level to top management. So, we
20:33will see two examples of the BI
20:36uh now. So, Neil, what are you showing
20:38us?
20:39>> What I'm showing here is
20:42um
20:43we've created a
20:45quite a lot or a few sets of reports.
20:48This set of reports we call it the CEO
20:51dashboard. We thought that this is the
20:53information that a CEO of a of a company
20:56would like to see, and it includes a lot
20:58of information in that case. I'm I want
21:01to show you, for example, the current
21:03production financial status. And we see,
21:06for example, that the highest cost site
21:09will be Aquamed. So, if I will highlight
21:11the Aquamed, of course, it will
21:14uh show me the information of Aquamed,
21:17but I can easily identify what's going
21:19on
21:20within my farm by sites, by unit groups,
21:24even by unit if I will call only A02.
21:28So, it will show me only what happens in
21:31A02.
21:33Uh this is more of a managerial, high
21:36managerial level report, but I can also
21:39go to a weekly report, which is more of
21:42an operational report, because it shows
21:45me what happened at my farm on a weekly
21:48basis. In that case, I took the example
21:50of um February to April 2025, okay? So,
21:54I can open it by weeks, but I can see,
21:57for example, here, this is a feeding
21:58report by week. If I will highlight only
22:01this week, I can easily identify what
22:04happened in this week across the
22:06different unit
22:08sites, sorry, but I can also go drill
22:11down all the way to the different units
22:14and understand where I had a problem,
22:16where I had deviations, where I had high
22:18deviation, small deviation, and so on
22:21and so forth. So, this is again a very
22:24powerful tool
22:25that allows us to easily identify
22:28interactively what happens at the farm.
22:32>> And again, what you saw is just two
22:33examples. The real power here is the
22:36semantic model where all the data and
22:38all the calculations of the KPIs exist.
22:41So, you know, you can access it creating
22:43your own create your own dashboards,
22:45create the management information system
22:48that is the dream of your company.
22:50You can access again the information
22:52with AI, etc.
22:54We are running out of time, so please
22:56could you please go to back to Aquamaze
22:58and
22:59so
23:01show two more things very quickly.
23:03And one is the redesigned planning
23:05functionality with a nice cool user
23:08interface. And we did that because we
23:10believe planning is a very important
23:13feature of AquaManager.
23:16Maybe so that we examine here like the
23:18results and
23:20and the harvest optimization.
23:24>> Yes, so we are going to results now and
23:25it will load, you know, as as you know
23:27again, the because all of you here or
23:30most of you AquaManager users, you know
23:32that a plan is a very detailed plan. So,
23:35now the system goes and calling a lot of
23:38information and when it comes back comes
23:42back again to the screen
23:43>> We compare plan to the reality, which is
23:45I think the main question. I mean, what
23:47was the plan, what we did? Let's make a
23:49comparison.
23:50>> Exactly.
23:51Exactly. And when we try to compare with
23:53lots of numbers, it's very hard to
23:55identify. We need to be number lovers.
23:58But, what we can see what we can do is
24:01go to visualization. And visualization
24:03allows us to easily identify where we
24:06are. Net growth plan versus
24:09reality. Fish number plan versus reality
24:11and so on and so forth.
24:13But, another very interesting thing that
24:17we can do now with the help of the
24:21of the AI is to optimize the um the
24:27cages, make a a um
24:29plan optimization
24:31>> the selection of the cages for
24:33harvesting.
24:34>> We want to select cages for harvest. So,
24:37we will go to the optimization bottom
24:40and we will type a query here. So, I
24:43need this amount of fish
24:46from that size and that amount of fish
24:48from this size and I give some
24:50restrictions and I send the answer to
24:53the
24:54AI agent. It goes back to all the data.
24:57It understand and knows the everything
25:00on each case that is
25:04dedicated to be harvested and brings me
25:06back the best option to harvest the
25:09cages with as less as possible remaining
25:13of fish in the cages and as less as
25:16possible
25:18or as high as possible accuracy for the
25:22order of the fish.
25:24>> Okay, we have only 1 minute. So, last
25:27but not least, talk about for 30 seconds
25:30about the smart unit alerts and then we
25:33go to the closing.
25:35So, can you do it in 30 seconds?
25:37>> the smart unit alerts, yes, yes, of
25:38course. The smart unit alert is
25:41second pair of eyes
25:43to make big eyes that allows you, farm
25:46managers, to know what happened in your
25:49farm yesterday or in the last period and
25:53it's a smart check of the
25:56of the farm, not just a regular check,
25:59that
26:00checks trends, checks everything that
26:03you need to know based on the
26:05configuration that you do, brings back a
26:07daily report on your desktop, which
26:11allows you to send the people that you
26:14need to the designated cages that might
26:18have issues today or identify the
26:22problems before it creates a big issue.
26:26This is the smart unit alert in a
26:27nutshell.
26:28>> Act
26:30be proactive and be fast. This is this
26:32is the thing.
26:33And thank you very much, Nir. Can we
26:35please go back to the presentation, the
26:37closing screen?
26:39And
26:40okay, what we tried to show you today is
26:43really the move from production records
26:46to connected control and the start of AI
26:50journey for Aqua Manager.
26:52This is not an idea for the future.
26:55The tools are here, the technologies are
26:57here, and we believe that companies who
27:01start early will learn faster, will
27:05adopt those technologies faster, and
27:08build a real advantage over over time.
27:12Uh
27:13really, we will be very happy to
27:15continue this discussion with each of
27:16you and explore how uh the new Aquamizer
27:21can
27:22can support you
27:23in a better way.
27:26We invite you to be among the first to
27:29take the next steps with us, especially
27:32in the AI.
27:34Uh we would like to thank you again for
27:36your time, your trust, and your
27:39confidence over the years.
27:41And a very big thank you
27:44for many of you
27:45for all the nice moments we shared
27:48together in
27:50social activities. Thank you very much
27:53again.
27:57>> Thank you also from my side, and I hope
27:59to see you around.
28:02>> We'll be in touch with all of you.
28:05>> Thank you. Have a nice time.
28:07>> Thank you. Bye-bye. Bye.

About this transcript

This page contains the full transcript of The Connected aquaManager Ecosystem: The next generation of production control - AI Launch Event by aquaManager, generated from the public captions YouTube serves with the video. The transcript has 3,795 words across 651 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.