UAS as a platform for integrated
sensing and ‘big data’
Prof Anthony Furness
Visiting Professor
Dr Tomas Norton, Senior Lecturer
Department of Agricultural Engineering
Unmanned Aerial Systems
in Precision Agriculture
Thursday 30th January 2014Agenda
The Green Revolution– origins of data developments in agriculture
Precision – the nature of precision
Precision Agriculture – defining precision agriculture Big Data – the features of big data
Big Data Analytics – the analytical counterpart of big data
Platforms for Data Acquisition – the available platforms and sources of data for agricultural big data development
Farming Data Hubs – potential role of farming data hubs in big data development
The UAS as a platform for data acquisition – specific attention to UAS as a platform for data acquisition
Big Data and extending UAS Capability – considering how big data can extend the capability of UASs
Importance of such Developments – importance in relation to national, European and Global needs
The Green Revolution
1940s scientific breakthroughs in plant genetics stimulated innovation in agricultural mechanisation, demonstrating the power of integrating
science and engineering for the benefit of society
So-called Green Revolution – this era not only provided precedence for innovations in food production but also demonstrated the power of data-driven decision support as farm extension services rapidly expanded.
Services exploited freely available Data (i.e. Open Data) in the form of weather and crop-growth forecasts to provide farmers with the capacity to benchmark and continuously improve in their capabilities to grow food.
(Benor, D., Harrison, J. Q., Baxter, M. (1984). Agricultural extension - the training and visit
system. A Worldbank publication)
More recent advances in computer science and technology during the past two decades has fuelled the development of Precision Agriculture
PRECISION
is about more precise (and accurate) measurement across all farming modalities and their use in developing more efficient and effective processes, practices and services, including the management of those process; it is about measuring,
understanding and reducing variability in processes, a total quality approach that embraces legacy and emergent
technologies to achieve the complementary goals of agricultural food production.
PRECISION
also implies more data, both from individual farming
applications and collective data sources from multiple sources The developments in the latter are resulting in very large data sources (Big data) that require new techniques to
Precision Agriculture
Wherein sensors, data processing and machine control enable the
conditioning of operations based on understanding the inherent variability in crop/animal production systems.
Precision Agriculture is now evolving into a paradigm where new knowledge in the biophysical sciences like photonics,
bio-electromagnetics and bio-fluidics are enabling the specific detection of physiological traits in crops and/or animals, intending to provide farmers with better opportunities to manage their systems.
Integration of data from different sources and levels, is becoming increasingly important, from commodity markets for price volatility
predictions to real-time weather, soil and air quality, and equipment usage for smarter decision support.
Greater data demands and opportunities are creating the need for
complementary data handling facilities and handling techniques – Big Data techniques
Big Data
Big Data
may be characterised by having extreme or variable values of one or more of the following features*:Volume (size of data set)
Variety (structure and range of data sets) Velocity (acquisition rate of data)
Veracity (uncertain quality or provenance of data)
Variability (in the meaning of data and relation to quality or robustness of data)
Complexity (with respect to relationships between data sets, sources of data
Demanding new approaches to maximise the value extractable from large and complex data sets.
*Big Data and Computing – Building a Vision for ARS Information Management, USDA Agricultural Research Service Workshop, Feb 2013
Big Data and Big Data Analytics
The Big Data approach requires less, but complementary dependence on the strictures of the causality-focused standard scientific method
The approach utilises vast quantities of data to achieve by-proxy
correlations that can assist in developing the foundations for Precision Agriculture
“Big Data Analytics”, how this approach is now termed, provides the potential to catalyse a new revolution in agricultural production,
presenting unprecedented opportunities for identifying associations
between information and knowledge entities, often faster and with greater temporal significance than conventional small data analytics.
Using the data from multitude of sensors embedded within fields, farm buildings, ground-machines, aerial vehicles and satellite platforms we can effectively inform predictive models that achieve insights and
Big Data Analytics
– extending the view
Big Data analytics this data can be reused, time and time again to reveal associations from different perspectives, such as context, intention,
objective, opportunities, constraints, know-how and so on, both within and across data sets.
With dynamic additions to data set content, data set types, data set merging, old data valuation and to the parsing algorithmic ‘windows’ maximum value may be extracted from the data acquired
The potential that this offers for agricultural development would appear to be immense, with parallels in service provision that are being seen for big data services in other areas of business activity.
Platforms for Data Acquisition
Global Navigation Satellite Systems
Satellite Remote Sensing and Imaging systems
Unmanned Aerial
Systems - Sensing and Imaging systems
Ground-based Sensing systems – fixed and mobile
Access to expert data / information, eg Evidence-based medicine and diagnostic services, Super-Navigator GPS-related services
information, (Human-machine-machine - Human) – Farming support
National / International survey data gathering eg Financial /economic/ resources / energy usage (Human-machine-machine - Human) - Farming support
Regional, National / International sensory data gathering eg For monitoring and protection of pooled resources / flood defences – (machine-machine - Human / Actuation) - Farming support
Regional, National / International data
gathering eg For forecasting purposes – weather/ natural disaster prevention (machine-machine-Human Activation) – Farming support Physical World Object-based Systems Internet-based facilities
Cloud computing – Object-based data processing needs (Human / machine-machine-Human) – Farming support
Remote data analysis, eg automated analytical services for industry, commerce and services eg. Mastitis detection (Precision livestock farming),
(Human-machine-machine - Human) – Farming support
Systems-defined automated software
downloads and up-dates for application and service systems eg. Security support systems, surveillance systems, transport management and associated information, mobile phone-based Apps – (machine-machine) – Farming support
Internet, Internet of Things
(IoT) and the ‘Cloud’
Farming Data Hubs
Big Data Providers
Data aggregation Data analytics Data services Data sharing
Farm-based Data Hub
Data tagging
Meta data tagging Data aggregation Data transfer Data Acquisition Platforms Farming Apps
Unmanned Aerial Systems for Data
Acquisition
UASs have gained a lot of interest in agriculture because they offer a range of attributes for remote data gathering, including:
• Near-real-time gathering of information from low altitude (< 120m) vantage points below cloud level (except fog conditions) on a whenever and where ever basis
• Low cost of investment and operation compared to common remote and proximal sensing systems;
• High potential for automation, which may enable inexperienced users to handle UASs with little training;
• Flexibility in choosing payload sensors and ground space resolution;
• Possibilities for use in actuator applications, such as synchronised mapping, cultivation, fertiliser application and pest control.
Unmanned Aerial Systems for Data
Acquisition
UAS platform hardware •Multicopter vs helicopter vs glider plane •Engines, battery •Sensors for geo-referencing (IMU, GNSS, ultra sonic) •Hardware design (protection) Sensor data processing •Radiometry •Geometry •Mosaicking •Storage, import, export•Meta data generation
Sensors for applications
•Multispectral
cameras (Vis, NIR, IR) •Spectrometers (Vis-NIR) •LIDAR UAS software •Autonomous navigation •Path planning •Obstacle avoidance •Sensor triggering •Autonomous starting/landing •Emergency strategies Applications of UAS Arable Farming •Plant production •Biomass mapping •Nitrogen estimation
•Water stress, irrigation
•Weed identification and control •Pathogen infection Livestock farming •Pasture management (biomass, quality) •Animal monitoring
(counting, weight, activity, health)
•Animal drive
Farm infrastructure inspection
•Roofs, solar panels,
irrigation systems, fish basins •Fences Actuators On board of UAS •Controllers •Sampler Other farm machinery •Fertilizer spreader •Sprayer •Sprinkler •Combine
Big Data Analytics to Support Farming
Operations
NG Os & Go v -so urc es of op en -d ata CIS CO- ork etw e n tiv va no in
in g to ols a nd se rv ic es SM Es - nd e a tiv va no in te ch no lo gic all y c ha lle ng in g se rv ic es
S
u p p o rt C en tr es : N C P F , S & W M CD
at a S ha rin g lo op s Farm International Plant/Animal Field/Barn Farm Regional Farm 1 Farm 2 Farm N Country 1 Country 2 Country N OpennessData, Tools, Networks
Continuous Improvement National Region 1 Region 2 Region N Continuous Improvement Continuous Improvement
Sensor data pooling
for
Precision Ag New Agri-Tech
Market data pooling
for Economic analysis Policy making objective driven context understanding
Regional data pooling
for Environmental management constraint awareness oppor-tunities know how based intention DATA
Big Data Extension
to UAS Capability
• Foundational developments in precision agriculture
• National, European and international statistics for farming development
• National, European and international standards for precision agriculture
• National policy and agencies for managing farming epidemics and pest control in farming
• Farming management and decision support in precision farming and input into achieving sustainable competitive agricultural economies
• Social inclusion in farming developments, including issues of food quality, food safety and national nutritional needs assessment
• Environmental management and factors impacting climate change
• Collective studies on genomics, proteomics and phenomics
• Evidential support in farming practice
• Developments in UAS systems and issues-handling in areas such as privacy, security of data and UAS governance
Big Data and big data analytics extending the capability of unmanned aerial systems by contributing to:
QuestUAS 200/300
Why is this important? - EU Priorities
• Business growth, research, innovation, enhancing ICT (including Internet of Things and Big Data impact upon precision
developments)
• Shifts towards low carbon economy (impact of precision on reducing carbon footprint)
• Environmental, Climate Change (Precision impacts)
• Employment & Skills (Precision impacts)
• Social inclusion (impact upon precision developments, including smart city urban farming)
All of these priorities are also embraced in a further priority – that is national, European and GLOBAL – Future food security
Thank you for your attention
http://www.harper-adams.ac.uk/initiatives/national-centre-precision-farming/