How Artificial Intelligence Will Radically Transform the Horticulture Industry
Horticulture is a data-rich environment with a lot of repetitive tasks being performed over & over again in order to obtain an optimal production cycle. It is therefore genuinly, fertile ground for analytics and artificial intelligence to blossom.
How will data & self-learning algorithms radically transform the horticulture industry? Read on to find out.
Why artificial intelligence will thrive in horticulture
Looking at the list of industries where artificial intelligence is likely to have the most impact, agriculture is often amongst the top-ranked entries. Such is especially the case for its sub-sector horticulture, which focusses on the cultivation of crops, often in an enclosed environment. There are a number of reasons for such a high score.
The cultivation of crops is a process that - as itself- generates a pool of interesting data. In order for flowers or plants to grow effortlessly, they require an optimal environment with given parameters as temperature, humidity, light intensity, certain levels of chemical substances, etc. With the availability of smaller sensors able to capture the data of such parameters, as well as the ability to connect these sensors in a network (the internet of things) and gather data centrally, a world full of opportunities emerges.
Add a wave of digitalization across the entire chain (raw material suppliers, growers, wholesale and retail) and one gasps at the long list of challenges that can be addressed by bringing together data from various sources, from various players in the chain. Data that can be used to fuel various AI-driven use cases.
Data that can be used to create insights and make better decisions, faster than ever before. And data that can be used to automate entire processes, driven by the rise in possibilities provided by a combination of artificial intelligence and robotics.
Let’s take a look at the longlist and identify the use cases where most companies will start to apply artificial intelligence, because of the business impact such is able to create.
The First Frontier: Using artificial intelligence to thrive in a saturated EU-market that is confronted with a lesser-predictable consumer
The limited opportunities for growth in European horticulture
Companies active in the European horticulture industry find themselves confronted with limited opportunities for growth.
Especially, when confronted with the current saturated European market. Looking at for instance the EU trade in fresh horticultural products, one is unable to deny the predominance of intra-European trade (compared to trade with non-EU countries). Moreover, the import of horticultural products in the EU has been growing faster than EU production. Numbers on for instance food & vegetables indicate that import is now almost four times the size of exports (7 billion euro vs 26,8 billion euro in 2017).
Add a stagnation of the volume consumed by EU consumers, changing trends and the impact of public concern with regard to food safety & horticulture's exploitation of natural resources, and one can clearly see the need for horticulturists to embrace change.
A number of strategies have been tested. Producers & distributors in horticulture have been looking towards the EU, to help strengthen the resilience of the sector (crisis prevention & risk management) and boost the consumption of their produce. The sector has been developing higher-value products, by either adjusting the product itself or by adding value through preparation, packaging or other.
Companies have attempted to obtain economies of scale through the acquisition of competing companies enabling the increase of for instance aerial size. Others have diversified their portfolio’s by acquiring innovative product concepts that can be marketed with more attractive margins.
Yet, the challenge remains and has become even more daunting due to the change in horticulture supply chain dynamics. The spectacular growth of the market share of national & international supermarket chains and an increasingly lesser-predictable end consumer have shifted the bargaining power.
Supermarket chains now call the shots when it comes to certain standards & procedures. They have pushed product liability and distribution risk more upstream, making such another reason for horticulturists to worry about their future.
Embracing the opportunities data, algorithms & AI present
As mentioned in the introduction, every single player in the horticulture value chain can benefit from gathering the data which the organisation generates. Be it for instance sensor data in the production environment or wholesale pricing data gathered over a given period of time.
Embracing the opportunities algorithms & AI present starts here. By crafting the appropriate data strategy and defining how the companies' data infrastructure will look like, to achieve pre-defined goals. It allows the horticulturist to gradually increase spending on setting up the appropriate infrastructure for each new use case it is willing to kickstart.
The first use case, with a lesser complexity, is the creation of crucial insights through dashboarding solutions which continuously deliver a comprehensive overview of business performance. These descriptive & diagnostic analytics enable decision-makers to better spot what is happening and why.
Looking at more advanced analytics, three use cases have proven to be genuinely effective in industries with similar characteristics. Let’s take a closer look at what they could mean for horticulturists.
Using data & self-learning algorithms to predict customer demand
As one continues to observe existing relationships in data, horticulturists can be equipped with predictive models able to provide information crucial for company decision-making.
These predictive models are developed using machine learning techniques which have proven themselves to be great candidates for such tasks.
A popular use case that is more & more applied in industries with similar characteristics, is the development of a demand forecasting algorithm.
Such an algorithm is designed to predict future demand for individual products or product clusters. Results of a particular algorithm help a company to streamline its business activities in terms of planning, asset allocation and business strategy.
Moreover, when a predictive model has gained an appropriate level of accuracy, horticulturist are be able to take a final leap and automate related businesses processes. Such would entail especially those business processes where a lot of data & information needs to be processed or where repetitive tasks are in play.
Using data & self-learning algorithms to set optimal price levels
In Horticulture, it has always been difficult to differentiate a product offer on other levels than price. Often, several attempts are made through marketing initiatives that try to justify a premium on the product. Examples include for instance Air So Pure, Decorum and Looye Honing Tomaten. Brands that aim to communicate specific premium characteristics in order to differentiate a homogeneous product.
As product price remains a dominant factor, insights in the pricing of competitors can be truly valuable. It can help a company to regain a competitive advantage, it once lost.
So, how can one cost-effectively gain such insights? Well, as a result of increased online sales through webshops or platforms such as Floriday, there is a similar increase in data available to learn from.
Price-scraping algorithms can collect the publicly-available pricing information and bring such together in a comprehensive overview. Machine Learning/AI algorithms will match the different products together, enabling a pricing manager to swiftly discover the price distribution for an individual product.
Such an overview allows innovative companies to develop a better-informed pricing strategy, permitting one to act fast on changes in the market.
Using data & self-learning algorithms to capture consumer trends
The horticulture & floriculture industry are no exception when it comes to trend sensitivity. Especially in the floriculture sector, trends have a big impact on market demand. But also fruits, vegetables and edible plants are subjective to consumer trends.
As the consumer becomes less predictable and requires a company to faster adopt changed needs & desires, easy access to trend information becomes a valuable asset for companies within the sector.
The growing trace of data points that end-consumers leave behind on the web, provides a tremendous potential for obtaining insights in consumer trends. Looking at how Koppert Cress used pictures from Instagram to detect where their edible cresses were used by top chefs from all over the world, it becomes clear that several players are thoroughly investing in order to obtain such capabilities.
AI-empowered algorithms will be able to extract information such as species, colour or other characteristics from pictures and videos, posted online. They will fuel the competitive advantage, innovative companies hold over their competitors.
The Second Frontier: Using artificial intelligence, but not merely to let robots to take over labour-intensive tasks
The horticulture industry is one with very specific labour characteristics. On the one hand, it has always been an industry that employed a lot of lesser-educated workers, working from season to season, performing repetitive tasks. On the other hand, there is the expert-grower that has gained years of experience on how to grow crops in the most optimal manner.
Using AI-driven robotics to take over labour-intensive tasks
Several initiatives have been taken with the goal to make horticulture less labour intensive, with artificial intelligence playing an important role in these innovations. They often constitute ‘the brain’ that enables a piece of robotics to perform the same task, highly accurate, time and time again.
Computer vision algorithms are for instance used, permitting robots to put young plants in soil. Or they enable drones to scan a given produce, looking for infected plants.
First attempts are currently undertaken to run an entire production cycle without any human intervention. A variety of sensors are placed in and around the produce, providing the datasets on which machine learning algorithms are trained to control the environmental parameters that impact the production.
A clear attempt to translate the existing expert knowledge into self-learning algorithms which would create a very scalable solution when proven successful.
Using artificial intelligence to make better business decisions
The majority of these aforementioned innovations is focussed on the production activities in a horticulture company. Yet, tremendous opportunities exist in adjacent business processes. It will not take long until advanced analytics and AI are used for tasks like planning, energy management, pricing and other business challenges.
Moreover, the wisdom on how to effectively run a horticulture company is slowly but steadily fading away. The average age of the grower is increasing and the pool of new growers ready to take over is limited.
Given the fact that artificial intelligence is so revolutionary as it learns from examples how to solve a challenge, it is no wonder that a lot of horticulture companies are putting there hopes on the technology to help solve the aformentioned issue.
Get started today!
In this article we have tried to shed light on the usecases that horticulturists will start to see popping up in their industry.
Demand prediction, dynamic pricing and consumer trend monitoring are concepts that will gradually emerge in the sector. Innovative entrepreneurs that embrace these algorithm-driven advanced analytics will take full advantage of the third wave in computing that is now flooding various industries.
They will come to terms with allowing data & algorithms to make decisions, once solely made by a human decision-maker. They will be the first able to take the full advantage of the growth in robotics and artificial intelligence to automate significant parts of their organization.
Finding the right balance between humans & machines making decisions and performing actions will provide these innovative horticulturists with a competitive advantage and provide a path for future growth.
Interested to learn more?
Guus van Heijningen & Nils Roelandt