What is the role of Bell Labs Solutions Research within Nokia and what is your mission as the president?
Nokia Bell Labs is the research arm of Nokia, founded in 1925 as Bell Telephone Laboratories, with a heritage of disruptive innovation in communications, networking, computing and foundational technologies. We have always had two dimensions to our mission: looking for disruptive ideas that address foundational problems of the times and solving these in the most impactful and meaningful ways. In other words, we are about forward thinking research placed in the industrial, real-world context. In addition to working on the direction of Nokia’s present portfolio of products, solutions and services, we have the ambition to constantly look into new areas where the company might wish to expand in the future. I am involved in that part of the research which looks at technologies that can advance the digitalization of industrial sectors beyond just the networking and connectivity aspects. That is my mission and it touches upon multiple vertical sectors, agriculture being one of them.
You partner with AeroFarms, a vertical farming company. To what extent do you believe that vertical farming can address the anticipated food scarcity problem?
I definitely think it will be part of the solution. Vertical farming has a number of benefits. One of them, for example, is that it can have many more growth cycles annually than traditional farming, while optimizing the use of water and nutrients and eliminating the need for pesticides and herbicides. There is also a strong sustainability aspect to vertical farming that comes from growing our food locally.
We work on a lot of AI, data analytics and machine learning solutions and when we met David Rosenberg, the CEO of AeroFarms, we realized that one of his main challenges was monitoring the vertical farm and understanding how it was performing, which is key when the growth cycles are that much shorter than in traditional outdoor farming. So, the problems that David articulated were very much in line with what our technologies could address, and this is how our collaboration began. Initially we thought it was a pure computer vision and data analytics problem. But then we introduced drones when we realized that we should address this at scale and be able to efficiently monitor an entire farm of tens of thousands of square feet. You do not need to monitor the plants all the time, but it is important to take a few images of every plant every day and drones allow for that by making the cameras and sensors mobile and bringing them to the plants to take images when needed.
To what other segments of agriculture can these methods, such as computer vision, be applicable?
I believe that all of these technologies can optimize food production in different settings. Anytime you need to collect data, these solutions can be applied to monitor and collect information across the entire supply chain. One starts with outdoor farming and vertical farming, but this technology can be utilized further up the supply chain, for instance, for food distribution (e.g. warehouse and inventory monitoring). So, beyond the more obvious usage of drones for precision agriculture, there are multiple other prospects.
‘Smart farm’ is becoming part of the global vernacular. Could you share your vision for the smart farm of the future?
The smart farm starts with collecting information and data on all aspects of the growth and farming processes and operations. Once we have this abundance of data, we need to extract knowledge and insights to understand the performance of the farm in real-time. Based on that, intelligent decisions can be made to optimize the production, increase yield, improve the quality of the produce, adjust the flow of nutrients or the lighting conditions and even perform predictive maintenance on the infrastructure of the farm to avoid any outages. The final step will be to tie that knowledge and the intelligent decisions into the actual operation of the plant and the control of the farm in a completely automated fashion. If, for example, we were to detect an area of the farm with poor growth, we should be able to use that information and have an automatic adjustment that fixes the issue. Then, the step after that, concerns the end of the growth cycle - automating the harvesting, the packaging, etc. I call this process of automation “sense, understand and act” which would run continuously. So, the smart farm of the future means being able to know everything that is taking place in the farm, and then having the corresponding decisions implemented in a productive, sustainable and safe fashion.
Speaking of safety, do you perceive the potential for cybersecurity risks along with full automation?
Yes, I believe there is a cybersecurity risk in all digitalized industries. At the same time, we are continually working on solutions to address that. The data being collected in smart farms is quite important and specific to the particular farm’s operation. This information needs to be protected from a competitive intelligence and business criticality perspective. We also need to ensure that the growth process cannot not be tampered with to ensure food safety and quality. One way in which you can ensure data security and sovereignty is by keeping the data local, on an edge compute and storage server inside the farm and not exposing it via the public cloud.
We are only starting to really understand how powerful AI could be. Do you see, like many others, potential pitfalls of embracing these solutions too fast?
We see many benefits of AI to both improve and optimize communication networks of the future but also to optimize future industrial processes. Nokia Bell Labs has a dedicated Responsible AI research team focused on ensuring that future AI systems are fair, accountable, secure and environmentally and socially sustainable.
If you had to synthesize a message to our readership about your work at Nokia Bell Labs, what would it be?
Our objective is that the technologies that we develop will find themselves across many vertical sectors, helping them improve their productivity, efficiency, safety and sustainability. We are, therefore, very excited to work with companies from different industries. In the case of agriculture, we are not agriculture specialists, but we understand the technologies that can benefit agriculture professionals. Innovation happens at the intersection of industries, and that is why we are ardent proponents of cross industry collaboration.
