To Sprint or not to Sprint?: The first leg of our Forest Fire project

Neshmiya A. Khan
Frontier Tech Hub
Published in
9 min readMay 30, 2022

As I begin working on this blog into the late hours of the night, the 27th of April 2022 to be exact, (it was also the holy month of Ramazan, which inadvertently turns us all into night owls) news started coming in that a massive fire had erupted in Margalla Hills National Park (MHNP), a national park situated in Pakistan’s capital city, Islamabad, and it was slowly increasing in size.

Incidents of forest fires, both man-made and natural, have started to become a common occurrence in Pakistan. Small campfire-esque blazes turn into raging infernos incensed by dry conditions, and further helped by rising temperatures, destroying already degraded habitats and killing endangered and threatened species of wildlife, such as leopards, fox, jackals, partridges, pheasants, and deer. Moreover, such incidents end up reversing whatever gains are made under the government-led large scale plantation projects, such as the Billion Tree Tsunami project.

Existing forest fire management systems are more reactive than proactive, and it is only once a fire has reached dangerous levels that action is taken to put it out, which often poses danger to the fire fighting team doing so and incurs far more losses that could be avoided with a system that anticipates fires and manage them in a timely manner. Hence, to address this issue, WWF-Pakistan sought to develop a forest fire management system with two main components: an early warning system to detect and warn local communities and the relevant Forest Department about forest fires as they arise, and a second system to model the spread of the fire to enable firefighters to plan mitigation measures, and more importantly inform them of where not go to ensure the minimum amount of harm.

What’s in a Sprint?

When we first came to learn that our proposal for developing an early warning system to detect forest fires was selected, it seemed an almost surreal moment. An idea that came about in random workplace discussions was now finally becoming reality.

When the process of contracting began, we assumed that this project would be implemented as any conservation project is implemented and executed. We follow the workplan and ensure timely implementation of our project activities.

However, in this instance, our assumption proved incorrect. We came to be introduced to the concept of “sprints”, to look at each activity as not simply a deliverable to be met, but rather as an “experiment”. Moreover, weekly conversations with the FTL team provided further insight into this new and, at least for us, unique process of not just project implementation, but we soon realized that this was also meant to aid in project design. Although we had already designed some high level activities, this allowed us to look at it from a more nuanced lens, and provided us with a space to further brainstorm these experiments. Albeit a bit confusing in the beginning, once we got a hold of it, it was kind of fun. More conversations led to each experiment becoming more fleshed out, and eventually our first Sprint was developed. It was finally time to put our experiments to the test.

Ready, Set, Go!

Our first Sprint was divided into 4 experiments to explore the viability of project implementation, 2 of which were quite straight forward. These included identifying a technical partner to design and develop the forest fire management system (which was easy as we were already working on a simultaneous project to develop another early warning system to detect leopards), and identifying and engaging a research consultant to conduct a study into forest fire incidents for the past 5 years (this would allow us to gather the data necessary to develop a robust forest fire management system). The latter simply required us to develop the requisite Terms of Reference and have them advertised through our internal processes.

Common leopard captured on AI Camera Trap [Copyright WWF-Pakistan and Lahore University of Management Sciences (LUMS)]
Common leopard chasing a porcupine [Copyright WWF-Pakistan and Lahore University of Management Sciences (LUMS)]

What we were most interested in within this first leg of the sprint was our interaction with the local communities in the proposed project sites, and the feedback from the relevant government stakeholders, specifically in the context of site selection, which we would use to gauge their interest and reaction to this proposed new technology.

However, these site visits proved to be much more fruitful and insightful than we anticipated, allowing us to identify gaps in our approaches which we had not even considered during project design (more below!).

Muhammad Waseem, Manager Conservation, WWF-Pakistan (and the brains behind the Forest Fire project) sitting with local community members [credit: WWF-Pakistan]
Consultations with the Forest Department, Khyber Pakhtunkhwa for site selection [credit: WWF-Pakistan]
Meeting with Forest Department, Government of Khyber Pakhtunkhwa, Pakistan [credit: WWF-Pakistan]
One potential site

The other aspects of the site visit were standard practice, but nonetheless still encouraging, such as receiving the full support of the Forest Department who even requested a formal presentation on the project, and then offered to provide office space in their building for the project team free of cost. These sites also have existing Community based organizations (CBOs) who were also on board. Having communities on board is vital as where there is community ownership over a project, there is a higher likelihood of its sustainability in the long-run.

Map of proposed pilot sites

What surprised us and why? And the challenges that came with it

This question is best approached in three overarching themes: “Sprint”, “Third Parties”, and “What we missed”:

  1. Sprint: The Sprint methodology was in itself surprising. The assumption was that we could go about implementing the project as we would usually expect to, but this method provided a new process which entailed us providing a basic concept to the FTL team, and then we would design and implement activities per Sprint on what was referred to as the “Pilot Log”. The Pilot Log asked us to identify the most critical and unknown assumptions that can be explored through experiments in order to (dis)prove our working hypothesis.

The challenge that came with this process was largely with time management. The concern was that the process of designing each sprint would eat away at the time meant to implement the project. And as the process was new, it did take some time to get the hang of it.

However, once we got the hang of it, it did get easy, and what made it interesting was how interactive it was. Providing a space for the project team to sit together and further discuss the concept with the FTL Hub allowed us to further flesh out the project idea, which made it into quite a fun learning and brainstorming exercise, and by building in time to reflect on the progress made on the pilot, the methodology enabled us to learn from our assumptions and modify the activity in light of the learnings

2. Third Parties: The forest fire early warning system as an idea had been brewing for a while, and we would often pitch it to potential donors and stakeholders in the hopes of being able to receive the requisite funding to work on it.

One such partner was Telenor, a telecom company with a strong presence in the north of Pakistan, where our project is going to be implemented. Telenor has a keen interest to work on projects that seek to conserve and protect biodiversity through new and innovative approaches, and discussions with their CSR team and their IoT department showed alignment in approaches and goals, especially in the context of testing and deploying new tech. It was only after a few discussions that Telenor agreed to support us in the implementation of this project, which included allowing us to use their telecom towers to deploy the early warning system, and even agreed to provide us with GSM connectivity as and when required in the future. An NDA is also in the process of being signed with the same, providing a strong indicator that the collaboration will be formalized.

Moreover, the reaction, interest, and ownership by the relevant government entities was also a good surprise. And this is important in terms of project sustainability. Getting third party stakeholders to be involved from the beginning is one of the key elements of project sustainability, as with interest comes ownership, and with ownership comes longevity.

3. What we missed: The site visit proved to be more than just a simple exercise in relaying information about the project and site selection, rather it allowed us to gain insights into areas which we may have not considered during project design. One aspect of the forest fire management that was picked up during these visits was what we intended to do once the fire was detected?

Albeit the spread software we hope to develop is intended to ensure that the firefighters are able to effectively plan how to put out the fire before it has spread significantly. However, by focusing on forest fire detection, we had not yet given much thought to the methods the firefighters would employ to put out the fire, and the protective gear they should wear. During our community engagement, we learnt that current methods of responding to forest fires are driven by community initiatives, such as community members cutting down tree branches to use as tools to put out fires, putting themselves in danger. In some areas, people did not receive proper training as to what to do once a fire erupts in the forest. Whilst this emphasized the need for a forest fire early warning system, these learnings informed us of our next steps and will be integrated into our second sprint in the hopes of making a more holistic forest fire management system. We are aware, however, that our proposed solution will not replace the forest fire management systems already in place until it is proven successful.

Moreover, another point we missed is that some fires are set by communities themselves to allegedly remove evidence of illegal logging. Trees provide a good source of fuel, and even income, and they tend to sell harvested timber for money, so how do we incentivize them to not do so? What potential alternative livelihood opportunities could be provided which could discourage such actions? This was another point we would now like to further delve into during our second Sprint.

What next?

As I write this conclusion (it’s 23rd May today) there is another story about a large forest fire raging in Zhob, Balochistan, decimating ancient pine trees and chilghoza forests. The fire started as a result of a lightning strike, and has been raging on for 15 days. There was initially no government action to control the fire, and the local tribal communities, who are reliant on the chilgoza (pine nuts) forests for their livelihood, attempted to put out the fire themselves. Three people from the local tribe however tragically perished in the fire while trying to put out the fire, while four others have been severely injured.

An emergency has finally been imposed by the Government, and Iran has also provided a special fire fighting plane to assist local efforts to control the fire. The news coming out of this incident is depressing — the tribes that reside in the region are reliant on the chilgoza (pine nut) forests for their livelihoods, and WWF-Pakistan has a close association with the area and why this news hits closer to home. Initiated in 1991, WWF-Pakistan’s project titled “Conservation of the Chilgoza Forest in Kasa Ghar’’ helped create a successful value-added market model for Chilghoza pine nuts which helped conserve around 9,000 hectares of vulnerable forests in the Suleiman Range of Pakistan and further sustained the well-being of forest-dependent communities of nine villages in the area. We now fear that those 9,000 hectares of forest may have been decimated, if not completely wiped out.

This incident lends urgency to the work we do, and underscores the necessity of the project itself. We look forward to Sprint 2 now, and are keen to inform the process of developing the early warning system with our findings in the first sprint so that incidents like this are prevented in the future.