Aquaculture Production Optimization through Enhanced Data Analytics - PowerPoint PPT Presentation

Aquaculture Production Optimization through Enhanced Data Analytics Aquaculture Open Data Cloud Innovation Joao Sarraipa, Kostas Seferis, Victor Prieto, Garry Cleere, Gary McManus, John McLaughlin, Tom Flynn, Ricardo Goncalves, Steven Davy Presented by: Joao Sarraipa UNINOVA

Why Data Analytics? » For instance: › Why a particular cage always has the fish that grow more efficiently? » “I think is because …” › If you u have e data to prove e your r stateme ements nts, you would say: » “It is because of …” » With increase of certainty - > new knowledge appears

How Data Analytics work? » Data analytics is the science of examining raw data with the purpose of drawing conclusions about that information [1] › To then reach to some conclusions that could end in new knowledge and consequent appropriate and effective decision making [1] Margaret Rouse (2016). Data Analytics Definition. In: A guide to HR analytics . Retrieved from the web at January 2016: http://searchdatamanagement.tec htarget.com/definition/data- analytics

H2020 ICT-15 15-2014: : 644715

AQUASMART Innovation Action » AQUASMART intends to solve a main problem that aquaculture companies are facing: › Companies cannot interpret the data they capture and also use the others data. » If they were able to do so, › they would be able to dramatically improve the production in terms of feed conversion rate (FCR), cost, mortality, diseases, environment impact, etc. .

Big Data and Open Data Analytics » Thus, AquaSmart aims › To bring Big and Open Data Analytics as a Service to the Aquaculture Industry › To create a cloud based platform with a backend based on machine learning and data mining techniques to provide assistance to aquaculture managers in the decision making process » Better view of the living inventory (biomass) that exist in a farm. » Be able to make accurate estimations of the growth of the fish.

AQUASMART Goal » The prime goal of AQUASMART is to accelerate innovation in Europe’s Aquaculture through: › technology transfer for the deployment of open data solutions › multilingual data collection and analytics solutions › turning the large volumes of heterogeneous aquaculture data that is distributed across the value chain, into an open cloud › Semantically interoperable data assets and knowledge.

Aquaculture Open Data Cloud Innovation App pplyin ing g Mode dels to Rea eal Dat ata

Example of Real Data

Normalised DataSets

DATASET MAPPING & VALIDATION » Input: Datasets from Excel files » Mapping the datasets with the semantics used in the AquaSmartData Tool › Sometimes new elements are introduced (private attributes) › Data types are defined to enable further transformations / validation ? FCR W C AquaSmartData Tool DATA DataSets Semantics private private 1 st step mappings FCR Water 2 nd step Temperature outliers identification

DATA REPRESENTATION: Interpolated Economic FCR (scatter plot)

DATA REPRESENTATION: Interpolated Economic FCR (surface plot)

DATA INTERPRETED: Bar plot of Relative FCR Error per test case (cages)

As-Is Scenario Ardag – pre new model

To-Be Scenario Ardag – with new model

AquaSmartData Training Analytics Programme

The AquaSmart training programme » AquaSmart provides ‘ An An analyt lytics ics too ool f l for or fis ish h farms ms ’ » To develop new skills, knowledge and competences in order to apply ly the AquaSma aSmart t Analyti lytics cs platf tform orm suitable table for fish h farm m produc ucti tion on to enhance production and efficiency.

What is the benefit of the AquaSmart training programme? Societal End-user Increased production Training for the AquaSmartData platform New business opportunities Develop skills and competences to apply data Knowledge transfer for sector analytics for enhanced production Supports standardisation Increased production Certification option (ECDL type validation) Increased sales Industry benefits Develop new business opportunities Workforce standard for sector Increased profits Colleges / Universities Increased proficiency European Commission Objectives Confidence in application of data analytics Blue Growth Policy Objectives Technical Training Knowledge transfer and training for effective use More educated decision making Optimisation for purchasing decision To support standardisation Certification option (ECDL type end-user validation) Inputs to software updates

Who is the AquaSmart training programme aimed at? 1. Business Owners 2. IT Managers 3. Farm Manager 4. Production Managers 5. Data Analysts

Training delivery modes » Tutor-led Training Certification › Traditional classroom › Virtual Classrom (Webinars) » Web based (e-learning + mobile) » Supported by: › Multi ti-language language options ions Analytics › Certif tificati ication on opti tion ons Gamification › Enhan ance ced d Knowledge ledge Transf ansfer er opti tions ns › Gamif ific icatio ation n options ions

The training courses » Course se 1: Concepts of Aquaculture Production » Course se 2: Essentials of Data Analytics » Course se 3: The AquaSmartData Solution » Course se 4: User Operational Features » Course se 5: Decision Making Support » Cour urse se 6: AquaSmartData System Integration » Course se 7: Industry Standards and Guidelines » Course 8: Business Dimension of AquaSmart in Aquaculture

Moodle Platform Aqua quaSm Smar art t LMS (Moodl odle e soon online ne)

Conclusions » AquaSmart Benefits for the Aquaculture Industry › Control the production process for maximum profitability, › Respond to a wide range of production challenges, in real time, › Identify, in a timely manner, production problems or trends, › Evaluate feed and fry suppliers, feeding practices and fish management strategies and › Continuously improve feeding and growth models.

The Current nt Mott otto Bringing Big and Open Data Analytics as a Service to the Aquaculture Industry

The Future ure Mott otto Bringing IoT to the Aquaculture Industry to enhance new knowledge acquisition and misperceptions prevention

AQUASMART Consortium The Technical Partners: The End Users: s:

The End Email: ail: info@aq @aquasm uasmar artd tdata ata.eu .eu URL: www.aq aquasm uasmar artd tdata ata.eu .eu Twitt itter: er: @AquaS aSmar martDat tData Link nkedIn edIn Gr Group: up: AquaSmar uaSmartData Data Facebook ebook Pa Page: e: www.f .fac aceb ebook ook.c .com om/Aquasmar asmartdata tdata Joao ao Sar arrai aipa pa jfss@unino @uninova.pt a.pt H2020 ICT-15 15-2014: : 644715