Innovation Design Engineering (MA/MSc)
Khulood Alawadi
Khulood Alawadi is an interdisciplinary designer passionate about working at the intersections of design with science, engineering and society and exploring tools for merging them. She is committed to recontextualising design for communities and environments that are seldom designed for, motivated by personal experiences and world events.
Previously, she worked as a designer, maker and communicator for science research institutes, cultural organisations and healthcare providers.
Her IDE group project, Fallback: Designing of Global Internet Shutdowns, has been featured in multiple publications including Dezeen, Creative Applications, ACM Interaction and awarded the Core77 Student Notable in the Design for Social Impact category. In the RCA x CERN Grand Challenge collaboration, her group project, KALA: A cross-cultural language game, was a finalist in the Social and Economic Disparities category.
The circumstances of the world during the development of this project emphasised the importance of adapting to new tools for prototyping and testing, recreating natural environments and conditions virtually and making engineering tools accessible to a broader audience and skillset.
This experience was also a lesson in empathy. It highlighted the disparities in human connectivity, accessibility to specific communities and environments as well as challenges with access to information and technologies to bring design and engineering concepts to life.
Aerial view of a date plantation
Inside a date plantation
The globalisation of trade has created significant growth in the date cultivation industry, rapidly increasing the number of new monoculture plantations which has unfortunately created an ideal ecological niche for biotic stresses to manifest. At the same time, the increased awareness of the risks of chemical pesticides and demand for organic produce encourages a new approach to agriculture, and more specifically, pest management.
Intervention Journey — The invasive Red Weevil has been the most dangerous and destructive pest to the date palm since it arrived in the region, eradicating entire plantations and years of hard work, causing not only environmental and financial damage but also aesthetic and emotional losses. The attack happens secretly and aggressively where visible symptoms of the infestation remain hidden until a very late stage, at which point the tree has low chances of survival. Infestations are not only very challenging to discover but are also time-intensive and physically demanding at the massive scale of these plantations.
The System — Palmspector is a robotic monitoring and inspection system and service for date palm plantations. This high-throughput system is deployed using a ground robot capable of autonomously navigating the plantation terrain to perform repetitive inspection tasks, systematically and consistently, to early and accurately detect infestations of the red weevil. Palmspector merges precision technology with traditional agricultural practices to augment the abilities of farmers by guiding them to the suspected trees. Palmspector is also a repository of knowledge on date palm health, filling the gap left by the loss of indigenous knowledge once transmitted through oral tradition.
This project applied design principles to develop a deep understanding of the context and pain points from the human and environmental perspective in order to be able to complement that with innovation in engineering and technology.
Technologies were validated through the use of computation and engineering simulation tools while the system sought feedback from key governmental and agricultural organizations, commercial farms, farmers and consumers. The validation showed great eagerness and acceptance towards autonomous technology alongside traditional agriculture, united by the common goal of better food and the protection of the limited natural resources.
Physics interactions
Simulation of the Palmspector system and tree approach protocol
Data acquisition from virtual sensors
Using simulated data from sensors to run algorithms in the virtual world
Real-time outdoor mapping
Point-cloud data, localisation and trajectory
This process focused on identifying the right technologies and designing the algorithms to navigate this complex environment (irregular terrain, vegetation, lighting conditions) that might vary from farm to farm or from season to season.
In addition to the real-life outdoor experiments, I built a simulator to create a realistic scenario of the date palm plantation to simulate the inspection process and tree approach protocol. This process included designing the virtual robot and programming the world; the palm trees, terrain and the physical attributes and interactions of each as well as integrating virtual sensors to react to the virtual environment. The algorithms designed on the simulator will be able to transfer directly to a real-world robot for evaluation and testing.
Prototype Sensor Fusion Data Collector
Prototype Sensor Fusion Data Collector
Prototype GUI
Capturing IR data
This prototype captures data from multiple dedicated inputs of the identified healthy and infested trees. This method acquires data points simultaneously following a uniform protocol and collates the output into organized datasets. These were then successfully used as training datasets for Supervised Deep Learning to detect a healthy or an infested tree