Weeds on hard surfaces do not grow at the same rate everywhere, nor are they assessed against the same quality standards in every location. While stricter quality requirements apply in city centres than in outer areas, priorities can even differ within a single area. This creates a key challenge for public space maintenance teams: where should you intervene, and where can it wait?
Especially at the start of the growing season, when all weed types emerge at once, this assessment and prioritisation becomes complex. Field teams must continuously determine where the risk of quality threshold breaches is highest. At the same time, it is becoming increasingly difficult to find enough skilled professionals who can make these decisions independently.
This is why there is a growing need to better support field teams with clear insight and prioritisation. In this blog, you will discover how AI-supported software helps make this knowledge available, enabling self-managing teams without specialist expertise to work more effectively and maintain better control over the quality of the public space.
At the beginning of the growing season, weeds on hard surfaces start to grow rapidly in many locations at once, often at maximum growth speed. The challenge is not only carrying out the work, but primarily making the right decisions: where is a quality threshold likely to be exceeded, and where can intervention wait?
Without clear insight, it becomes difficult to deploy teams effectively. Teams must estimate priorities themselves, while in practice this knowledge is becoming increasingly scarce. As a result, organisations often end up treating the entire area out of necessity, while the real need is to focus on locations with the highest risk of quality breaches. This leads to fragmented work, increased operational effort and a higher risk of non-compliance.
The core challenge is therefore not the execution of the work itself, but the continuous prioritisation of it, without requiring additional skilled capacity in the field.
AI-based image recognition makes it possible to fully automate the identification and prioritisation of work. Cameras mounted on vehicles capture images of the public space at fixed intervals. A trained detection model analyses these images and identifies where weeds on hard surfaces are present and where quality thresholds are at risk of being exceeded.
This information can be delivered to field teams in two ways. In the first approach, a map is presented showing exactly where action is required and where it is not. In the second approach, tasks are automatically generated based on these insights, providing teams with clear, actionable instructions.
This creates a clear way of working for field teams: they know exactly where to go and what needs to be done. It enables teams to operate independently and continuously set the right priorities.
The result is less dependency on individual expertise, more consistency in execution and better control over the quality of the public space.
With Jewel Software, weed control on hard surfaces shifts from reactive to proactive. Instead of only providing high-frequency insight, the system translates this insight directly into execution proposals and priorities. Automatically generated tasks clearly indicate where action is required and which locations should be prioritised.
Jewel continuously tracks which tasks have been completed and which are still pending. By combining historical data with real-time insights, recurring problem locations become visible, enabling more targeted interventions. In practice, this often shows that parts of the work area do not require treatment, allowing teams to work more efficiently and avoid unnecessary effort.
However, the real strength lies not only in the technology, but in how it empowers teams.
Within Jewel Grounds Maintenance, the focus is on supporting field teams in making the right decisions. By providing insight and prioritisation directly within the execution process, teams can work more independently, collaborate more effectively and respond to daily dynamics.
This results not only in a more efficient operation, but also in better control over maintaining a clean, safe and liveable public environment, every single day.
Organisations working with this approach gain continuous, high-frequency insight into the public space and can steer more effectively on quality. The risk of non-compliance decreases, operations become more efficient, and even teams without specialist expertise can perform at a high level. At the same time, clients benefit from better-informed decisions and measurable results.
Beyond weed control on hard surfaces, image recognition is increasingly applied across other public space maintenance processes, such as street sweeping, side waste near underground containers, leaf and blossom debris, and litter. High-frequency visual data enables organisations to manage multiple processes within a single approach.
Would you like to benefit from this way of working? Get in touch and discover what this can mean for your organisation.