Unlocking the value of data in outdoor agriculture
31 January 2022
While technological innovations have abounded for several years in outdoor agriculture, why most of them struggle to enter the fields?
Are you struggling on what will be the next digital transformation of your company? Well, it may be becoming a “bricks & software” one !
During the first Internet revolution, around the year 2000, traditional “Brick & Mortar” companies had to gradually carry out their digital transformation, becoming “Click & Mortar” as the expression goes.
First a showcase website, then a merchant website, eventually some interfaces with customers and suppliers (extranet, EDI, API…). This revolution was often led by dedicated teams, often first from the communication department, then from the digital department, and finally involving the relevant departments (sales, logistics…).
All of this was done by adding skills and adapting processes, sometimes at a forced pace driven by pure players, but without challenging the company’s know-how.
At the same time, processes were being digitized, continuing the evolution of the last century towards office automation, ERP, computer-aided design and management tools, robotization and automation, and then adding sensors and data management tools (Factory 4.0).
Often, the tools were bought, more or less on the shelf, and adapted, integrated, generally by consultants, to fit the processes of the company, which remained master of its know-how.
Software, digital, and even data were at the company’s borders. Only a few companies had already integrated software as their core since a long time, the few exceptions being in finance, and of course the pure software companies, and some fabless companies that succeeded thanks to their mastery and competitive advantage on software (e. g. Apple).
Today, we are witnessing a more profound revolution, which raises questions about the evolution of the company’s core business and may put the mastery of its know-how at risk.
Indeed, software is increasingly capturing the intelligence of the company’s processes. Software is no longer the tool (transactional ERP, more or less commoditized), it is the process. Artificial intelligence, in particular, has managed to capture so much know-how that it is becoming risky to outsource it. Specialized startups or large software companies are trying to capture the data of traditional players to train algorithms. If the symbiosis is tempting, the risks of value capture are high. In the industrial world, the risk is similar to an uberization, where the value (design, planning, maintenance, etc.) is concentrated in the software, the industrial player becoming a simple executor. In services, the risk is to be purely and simply replaced by a one hundred percent dematerialized service, or in any case by one where the human being is confined to standardized, interchangeable roles.
On the other hand, for hardware companies (from cars to electronic devices, home automation, etc.) as well as for service companies (banking, platforms, etc.), software is the customer’s main interface with the product. Even though most of them have understood this, they are struggling to bring their interfaces and software production processes up to the state of the art. Salespeople and consultants are no longer on the front lines, but interfaces are. The good news is that it may be easier to make great software than to recruit and train a team.
In the era of software companies, or rather “Brick & Software”, the decisions to make or to buy, and the management of priorities, must be completely revised, giving a greater share to the internalization of the manufacture and control of the software, and in particular of artificial intelligence and ergonomics, while raising the quality to the state of the art of pure players.
So how to go about it?
This shift is more complicated than the first digital transformation, which could remain relatively at the company’s borders, or even be outsourced. The focus must therefore be on three areas: human resources and skills management, organization and culture, technology and infrastructure.
How do I recruit, do I need a new employer brand? Can I integrate these new talents into my traditional career path? Should I centralize a team to have a critical mass or on the contrary, spread them out as close to the field as possible? How can I get help? How to industrialize, on which infrastructure, how to articulate the software with my installed IT?
There is no recipe adapted to all companies and their organizations. Nevertheless, certain invariants and major options are emerging.
Regarding human resources and skills management, we have seen that it is often possible to upskill existing employees and recruit new talent without necessarily revolutionizing everything. However, a visionary and wise impulse from the CTO or CDO is necessary to initiate the change, to attract and guide. A major training effort will have to be deployed to enable the upskilling.
The organization and the culture will be key to allow these talents to deliver their added value. Even in the age of StackOverflow, GitHub, and other platforms, a critical mass is needed to form a community that will help and complement each other, and offer inspiring career paths. The community does not necessarily mean a centralized team. It can be spread out near the field, but it must be strong enough to meet the requirements mentioned above. The culture will have to accompany this evolution: flexibility of processes (HR, IT, etc.), agility, best practices and state of the art of pure players, and shared by the entire hierarchy.
An effective practice to succeed in this new digital transformation is to have a central team at the forefront, incubating complex projects and disseminating them to the more field-based teams. As long as you identify high-value projects and stay in touch with the business, this organization allows you to innovate and deploy solutions that work.
The technology and infrastructure will have to follow: open platforms allowing continuous integration and delivery (CI/CD), access to open resources (web, open source, open data, API, etc.), modularity of the architecture (serverless, etc.), scaling, etc. These flexible platforms will be able to interface via APIs with installed IT systems. It should be noted that the emergence of these platforms, architecture, practices and resources has considerably facilitated the internal recovery of the software, as developers and designers can focus on the business part and not on the lower layers.
When scaling up, it is important to develop the various components (use cases, infrastructure, governance, etc.) gradually and in parallel to avoid failures, delays or investments with too low a return.
The most efficient organization is to have teams that integrate exploration, design and operation (ML Ops / Dev Ops). Nevertheless, a handover to the IT teams is possible as long as the culture is close to the IT teams, as long as they think in terms of industrialization by design (no rewriting of code!), and as long as they have short and efficient feedback loops.
Of course, these main principles must be adapted according to the size, organization and culture of the company, and the digital transformation project towards the “Brick & Software” state of the art will have to make the different teams become more skilled, while enhancing the know-how and culture of the company.
Sur le même sujet