The Final Sprint in Distributed Manufacturing: Conducting Hyperlocal experiments

Kuldeep Bandhu Aryal, Frontier Technology Specialist at BRAC Social Innovation Lab

Before going into the details of how the sprint 4 activities were conducted, the first blog will give you an overview on the motivations behind the sprint 4

Demand Aggregation: Exploring existing and established platforms

As we started the sprint 4 experiments, we also explored existing platforms that could allow us to aggregate real time demand for WASH products from the field. We explored both modifying existing open source platforms like Kobo and other established platforms that were already ahead in terms of matching manufacturers with demand like and Needslist. Needslist was also used as a demand aggregator and matching platform for providing access to PPE for frontline workers in a HGC funded project in partnership with Field Ready and OSM.

The challenge however remains in linking potentially open source solutions like the DM platform to existing platforms that are not fully open. Unfortunately, this creates silos amongst solutions which if linked together could have provided a critical breakthrough for adoption of new innovations in the sector. The upcoming road map for the development of DM platform will have experiments with both open source as well as closed source platforms to understand in practicality and efficiency gains from either of the platform.

Hyperlocal Distributed Manufacturing: Potential for impact and challenges around it

As we completed 3 sprints over the course of the grant, we were able to develop a system architecture for the DM platform that allowed, in theory, a combination of principal and secondary manufacturers to engage with BRAC. The important part of the system was the quality control (QC) mechanism that allowed BRAC or an independent QC party to have oversight over the quality of the product manufactured.

The previous sprints explored that feasibility in Dhaka, but as we explored the possibility of increasing the impact of the DM and providing the most value to both the manufacturers and humanitarian agencies, we realised the need for a mechanism that allows hyperlocal DM in Cox’s Bazar itself.

In the 4th sprint, we conducted a hyperlocal experiment of DM where we trained 21 local female manufacturers in the host community. We partnered with the HCMP WASH team to conduct a two day training where we trained the participants to manufacture soap and handrub. The experiment had to go into this direction in order to align the 4th sprint with what the HCMP programme categorised as an ‘innovation’ activity in their project that was funded by another donor. This meant that HCMP could invest their own resources during the experiment and on the other hand the DM platform will then also gain mainstream recognition as an innovation process that was used by HCMP programme. This has lead to increased ownership of the DM process into HCMP as well as given confidence to the programme in their ability to manufacture high quality aid product without the need for engaging with large scale companies in Dhaka.

The HCMP programme supported in the procurement of training materials, equipment for manufacture while the project team provided technical assistance by providing technical training on manufacturing technique, quality control and access to maker manual and calculation techniques for preparing the Bill of Materials (BOM). The aim of the HCMP programme was to provide income generating skills to the host community through entrepreneurship development.

However, one of the big challenges that emerged was that even though the BRAC HCMP programme provided training and capacity building to female ‘micro manufacturers’, we did not yet have a mechanism established to incorporate them and buy their manufactured products via the procurement system. This revealed a challenge to the existing DM platform that envisions Principal manufacturers as legal entities for engagement with BRAC. The micro manufacturers did not have the entire set of legal documentation (TIN, trade licence and BIN) to engage with BRAC and the cost of supporting them to formalise was not incorporated in the project budget and the timeline.

Engaging with informal manufacturing sector

The informal manufacturing sector in Bangladesh, is a major employer for low and middle income people in the country. The sector has deep linkages and relationships with the private sector, taking even sub contracts of some items or working in ‘spare and repair’ processes. However, humanitarian and development agencies cannot engage with them in any meaningful way due to the high threshold of legal due diligence and documentation for both vendors and partners. One of the major learnings we had around meeting local demand with local supply is that humanitarian agencies approve in principle the value and impact of DM. The stakeholders were mostly interested in the social and environmental impact of DM based on the impact analysis conducted in sprint 3. However, we were also able to identify institutional constraints in the system that limits the ability of actors to act on it.

Thus, unfortunately, we as a sector that has a mandate to work with the marginalised and support the informal sector, are ourselves not able to engage with them. As the 4th sprint put this issue front and centre to the leadership at the HCMP programme, the HCMP WASH unit expressed commitment to engage with the informal manufacturing sector, especially if the manufacturer is operating inside Cox’s Bazaar by the end of December, 2022. This will help simplify the DM architecture by removing additional barriers on the dual layer of manufacturers, but will require additional focus on quality control and increase in the Acceptable Quality Level (AQL) for the local manufacturers.

Integration with supply chain system

Another lesson that emerged was on linking the manufacturing partners with the supply chain for procuring raw materials. The majority of supply chain actors in Bangladesh are based in Old Dhaka and the local manufacturers in satellite cities like Cox’s Bazaar have to be dependent on the middle men for access to even raw materials. We shared the list and map of raw material suppliers to both local manufacturers and HCMP wash for due diligence and getting fairer prices for the products in need. This is another big gap in creation of the DM sector that we have to fill to ensure fairer trade practices in the supply chain and overall reduction of cost and increase in quality of the products that will be manufactured using the DM platform.

Production Contract Challenges

The initial conversation around experimentation with the HCMP WASH was around undergoing a traditional product procurement process of BRAC. The WASH team also proposed an initial list of local WASH manufacturers that have previously engaged with other local NGOs in locally producing WASH items and were confident in their ability to provide a production contract to the manufacturers. The project team also provided continuous support and justification via HCMP WASH to undergo the production contracting process for the selected manufacturers.

In addition to sensitising WASH and the HCMP procurement team in the two day technical training we conducted, we identified and selected 7 high performing candidates to conduct the DM experiment.

As we went through with the process of production contract, challenges and issues were raised around the selection process of the vendors as well as the legal threshold that the manufacturers have had to pass through to engage in a tender process at BRAC. During the first DM experiment we conducted in the second sprint, the team had to make a single source justification with the Chief Financial Officer (CFO) of BRAC to select and identify the manufacturers for the DM experiment. But because of policy changes in the HCMP programme, the SIL as well as WASH team was not able to pass through the justification process required for the selection of the vendors.

The HCMP team in this experiment were comfortable with purchasing the raw materials by themselves and paying the makers additional amounts to cover their labor, packaging, quality control and profits instead of paying for the final product from the manufacturers. Under this condition, the DM architecture can allow organisations to directly work with secondary manufacturers especially in a case where the manufacturing infrastructure is centralised but manufacturers act independently during the production process. This is not an ideal system and has visible flaws around the business modality and overall sustainability of the system. Under this model, the whole process has to go through single source justification and will require a lot of push from motivated individuals inside the programme. However, it can be taken as a precursor to a more independent DM system for some humanitarian agencies.

The Shift

Thus the team had to shift the modality of the experiment where the BRAC HCMP programme would procure the raw materials and provide expenses incurred for the local manufacturer around labour, packaging, quality control and profit margins. This then led to a delay of procurement where BRAC had to undertake a requisition process for raw materials and tools for the manufacturers.

As we completed the 4th sprint, there were primarily three major learnings that we can summarise from the background we have presented above:

1. There is a clear need for making policy changes in the procurement process and reducing threshold for local manufacturers as a viable and at times preferable option for manufacturing aid supplies
2. Based on the impact mapping analysis and engagement with the stakeholders including local manufacturers and BRAC field staff, hyperlocal DM has the most potential to create the most impact as we move ahead with localisation of the Rohingya response.
3. With the BRAC HCMP programme understanding the challenges and barriers for implementing DM in Cox’s Bazar, they have shown interest and commitment to make policy changes to their procurement process.

Thus, in order to conclude the efforts of this project and livestream this radical and disruptive system into one of the largest humanitarian actors in the region, we now need a sustained push on both policy changes by the end of the year as well as continue to modify the DM system that we have built in this project.

Link for the Part 1 of the two part blog is here.