Skip to main content

New intelligence in the field

Robotics and swarm technology

Larger working widths, more powerful engines or greater throughput - quantitative growth was and is an important aspect in (agricultural) technology. And even if not everyone admits it: Besides the usefulness, it's also a matter of appreciation of the sheer size. Thomas Herlitzius and his team are taking a different route. "Field swarm" is the interesting sounding name for the technology being researched by an interdisciplinary consortium. Herlitzius, Professor of Agricultural System Technology at the Technical University in Dresden, Germany, is the spokesman for the collective consisting of the university, business partners and the Fraunhofer Institute for Traffic and Infrastructure Systems (IVI) in Dresden.   

Departure from maximum individual capacity 

The principle seems plausible: Instead of one large machine, several flexible, combined units are used. The field swarm consists of automated and self-propelled units. For example, the grubber is no longer a single unit with a nine metre working width. Instead, several smaller units are (virtually) interconnected. Depending on the working field size, it is not necessary to provide the maximum capacity every time; it can be less. That protects the soil and saves energy. And alternative drive energy is planned for the field swarm anyway. Agricultural system technician Herlitzius, a pragmatic visionary, and his team warn against unrealistic, i.e. premature, expectations. 

"Triggered by a technological innovation, first comes the peak of unrealistic expectations, followed by the valley of disappointments", says Herlitzius citing the "Hype-Cycle model" of the US market research company Gartner. Ideally, this is then followed by the path to enlightenment, which leads to the plateau of productivity. 

Capacity utilisation as target 

To put it less metaphorically: Perseverance is required. The research of the consortium coordinated by the Technical University in Dresden is directed at capacity utilisation of the existing capacities instead of individual top performance. Farmers and consumers benefit with robotics as the highest level of automation: Producers with resource-conserving effort and consumers, as the new technology enables an extremely reliable tracking of raw materials in foodstuffs. 
Automated processes suitable for use in practice that go beyond the swarms, including steering systems, have reached the "plateau of productivity". However, in agricultural soil tillage we are just beginning, explains Herlitzius and introduces a complicated technical term to the discussion: "Collaborative Robotics", or Cobotics for short. Humans and machines move more closely together again here. Machines support and correct human work. For example, in industry a gripper arm gives the technician at the assembly line the parts to be installed and nudges the worker if necessary to avoid injuries. Collaborative robotics suitable for use in practice doesn't exist yet in agriculture. 

Initial approaches are assistance systems in which human beings almost become a "sensor", for example in combine harvester adjustment, and confirm decisions ("headland management turn").  Scientists evaluate this as "hardly variable and always with a certain context". Real collaboration adapts to the respective current situation and the competence of the human being. Until this is possible, a great deal still has to happen in the development departments of the agricultural machinery manufacturers.

Urgent need to clarify data ownership 

When asked about the options for artificial intelligence (AI) in agriculture, Herlitzius rejects its usual definition as self-learning information technology as too far-reaching: "Artificial intelligence is primarily based on statistical methods and not - as is often depicted - on an independent world of thought". The latter continues to be reserved for human beings.  
Artificial intelligence is mainly a question of computer capacities with which the options can be checked extremely quickly.  In addition to computer capacity, statistical methods of AI also required a broad, high-quality data basis. And that is lacking in agricultural technology, explains Herlitzius and adds as spokesman for the scientific consortium: "We are just beginning to collect and therefore are having a serious discussion on the control over data. He says there is a clear answer to data sovereignty: "It must lie with the farmer or the machine owner." 

Progress occasionally begins easily in a literal sense: According to Herlitzius, the Field Robot Event during the DLG-Feldtage (Field Days) is excellently suited for testing navigation of agricultural machines on a small scale. The scientist from Dresden attaches great importance to the entire research of his team and its partners being directed at promoting sustainability in agriculture - from efficient use of natural and operating resources to a transparent value-added chain to more biodiversity. A thought that responsible farmers the world over are likely to share and that is reflected in the guiding theme of AGRITECHNICA 2019: "Global Farming - Local Responsibility". 

At a glance

  • Field swarm: Flexible smaller units instead of large individual capacities
  • Cobotics (Collaborative Robotics) connects humans beings and machines 
  • Artificial intelligence is primarily based on computer capacities 
  • Automated soil tillage is still in its infancy 
  • Sustainability as objective of field swarm, cobotics and digitalisation  

Weitere Beiträge zum Leitthema "Global Farming - Local Responsibility":