(1/3) How to Test if Your Organisation is Ready for Distributed Manufacturing?
A step-wise guide to design experiments to explore the feasibility of aligning a distributed manufacturing system with procurement processes in organisations.
Based on insights derived from BRAC’s experience in exploring the scope, challenges, and opportunities in mainstreaming distributed manufacturing in South Asia
*Written by Chathuri Weerasinghe, Social Innovation Lab, BRAC*
_________________________________________________________________________________
The Context
Following the reflections made in the last blog, our efforts in exploring the potential feasibility of distributed manufacturing in South Asia continued to grow and expand within a time span of 3 months. Building on the initial learnings of the exploration- which brought to light the potential BRAC has in mainstreaming massive-small manufacturing in Bangladesh, as well as the unique position it is placed in to enable a prospective sustainable market for local manufacturing within and beyond Bangladesh- we focused on designing and executing a ground experiment, which offered us the opportunity to test the feasibility of a distributed manufacturing system within a context such as Bangladesh.
To better understand the contextual requirements for mainstreaming distributed manufacturing, the test was designed around the BRAC aid supply system, representing the development and the humanitarian sectors, and with a selected sample of makerspaces and Fab Labs, representing the makers ecosystem in Bangladesh. Hence, the experiment was designed not only with the purpose of exploring the potential in establishing a customised and a contextualised distributed manufacturing system within BRAC, but also to examine how such a systems-change initiative can leverage the capacity of the involved manufacturers, as well as pave the way to enabling more effective real-time on-the-ground BRAC responses.
Why Experiment? Why not Just Implement?
In reality, establishing a distributed manufacturing network would, in general, only entail connecting geographically dispersed manufacturers and the demand generation units via a digital platform, which puts the system in charge of the rest of the work. But, aligning such a system with a robust institution can be more complex than that especially due to the following reasons:
- There can be hidden challenges we do not account for at the beginning, which usually do not become an issue until it comes to the point of implementation. An experiment can lead to opening up these hidden challenges, by stimulating extra scrutiny stemming from the added sense of accountability among the parties involved in the implementation process.
- This means that there is room to identify these challenges at an earlier stage, instead of having to experience surprises all along while trying to implement a distributed manufacturing system in an institution.
- Experimentation creates room for failure and helps the implementers do in-depth observations along the process, and find alternative pathways for the planned ones that do not initially work out. Whereas in implementation, you are often expected to make things happen, which is not always the most possible thing, is it?
- Experiments help us understand the strengths and weaknesses of the actors involved, and how they act under changes, stressed and challenging conditions. This helps the implementers map the stakeholders ahead of time, giving them more time to familiarise the stakeholders with the concept, the system, and the process, as well as to use their strengths to the better advantage of the system implementation.
How to Design a Distributed Manufacturing Experiment?
To enable and encourage more experiments that can guide organisations to explore the feasibility of aligning a distributed manufacturing system with their supply chain and procurement units, we compiled our experiment analysis into a 14-step guide that is replicable and customisable in any context.
Before understanding the steps that lead to an effective distributed manufacturing experiment, it is important to know in which phase each step becomes the most effective and contributes to increasing the overall efficiency of the process. Based on our analysis, we observed that the experiment design requires the completion of three phases: Plan, Prepare, Launch.
Phase 1: Plan
The planning stage is critical, and will require the completion of the first four steps in the distributed manufacturing experiment design.
Step 1: Know your target user-base
A distributed manufacturing system runs as a demand-driven manufacturing process, which puts a tremendous emphasis on the user preferences and user needs over organisational procurement requirements and manufacturer-activated supply flows. Hence, involving the target user-base in the planning process is preliminary to designing a distributed manufacturing experiment for the following reasons:
- Helps understand the types and quantities of products required, and at which level of urgency
- To understand the requirements for customised designs that best fit the users’ context, need and use case
- Helps understand the willingness and positive outlook in providing a market for local manufacturers
- Enables and encourages more interest in generating real-time demand
Example:
For the current experiment, BRAC chose its Humanitarian Crisis Management Programme (HCMP) as the target user base. The HCMP currently operates in Cox’s Bazar, Bangladesh, with its interventions and services targeted at supporting the Rohingya and the host communities in the region- representative of a human service delivery system. While there is no set rationale behind choosing the target user-base for a distributed manufacturing experiment, we chose this particular stakeholder so as to get comparative insights into which means a distributed manufacturing system can bring in positive changes into a procurement system within an emergency response context, as well as within a non-emergency response context.
As one of the first steps in the experiment design process, the BRAC team visited the HCMP Cox’s Bazar office for one-on-one consultations with its sector leads (I.e., WASH, Health & Nutrition, Education, Protection & Child Protection, Agriculture) to better understand their product needs and requests for customisation. The consultations resulted in the successful outcome of a product portfolio that not only was representative of the user needs, but also was practical in terms of the selected manufacturers’ production capacities.
The finalised product portfolio included images of the products to be manufactured (I.e., 3D printed face shields, WHO approved hand rubs, Foot Operated Taps, 3D printed Umbilical Cord Clamps, 3D printed ear guards for face masks) the Bills of Materials (BOM), uses of and the applicable user base for each product.
Tips:
Remember, it is an experiment! Hence;
- Keep the product portfolio limited to more simple designs and limited complex designs
- Set realistic targets for demand points that are to be collected
- Make use of open-source designs
Step 2: Stakeholder mapping and understanding system dynamics
Once the user base of the experiment is identified and the justification for doing so is clear, the next step would require understanding the rest of the stakeholders that will be involved throughout the experiment, as well as the type of contributions they will be making to increase the effectiveness of the distributed manufacturing platform. The stakeholder and system dynamics mapping can be done following the below framework.
Example:
In the case of BRAC’s experiment, the stakeholders were mapped under the following categories upon understanding BRAC Aid System’s supply chain:
- Procurement
- Demand generation
- Supply and quality control
- Product distribution
Step 3: Building testable assumptions
What is an experiment if it is not developed on the foundations of testable, and falsifiable hypotheses? Hence, it is important that the planning phase of the distributed manufacturing experiment also focuses on building strong assumptions/ hypotheses to be tested along the experiment.
Tips for building strong assumptions for a distributed manufacturing experiment:
- Play the devil’s advocate: Assume everything will fail and all stakeholders will be detractors so that you can imagine different possibilities
- Research on what has been done before and understand what already exists
- Understand how the current system (non distributed manufacturing) works and find verifiable strengths and weaknesses of the system
- Note the most uncertain and “lethal” assumption of the work you will be doing: This can walk you to achieve big gains quickly in terms of learning
To give more clarity to the assumptions and turn them into a framework that would guide the framework of activities (Step 4), answering the following questions under each assumption will be useful:
- How will you verify this assumption?
- What is the minimum proof you need to verify this assumption?
Example:
BRAC’s distributed manufacturing experiment was designed around the following overarching hypothesis, and six assumptions that further narrowed down the hypothesis into more specific testable and falsifiable conditions:
Step 4: Designing the framework of activities
Now that the assumptions to be tested are developed, it is time to draft the activities to be completed in the experiment!
For this step, going back to the stakeholder mapping will be useful. Consider the stakeholders that were identified along the supply chain where the distributed manufacturing system will be introduced, and brainstorm the activities that constitute the experiment around the stakeholders and the assumptions developed in Step 3. Questions to answer while mapping the activities entailing the experiment can be:
- What are the things (I.e., documents, information) that we need from each stakeholder related to the distributed manufacturing experiment, and how do we collect them?
- What are the tasks that need to be completed by the stakeholder?
- By when do we need this information or tasks completed?
- Who will be in charge of ensuring the deadlines are met?
- How will we track the progress of completing these activities?
Example:
BRAC’s framework of activities was designed around the four main categories of procurement, demand generation, supply and quality control, and product distribution as well as around the stakeholders that were identified along the supply chain of the BRAC aid system. The plan encompassed the tasks to be completed, with deadlines and assigned focal personnel from the experiment team, along with an option to update task progress.
