Q&A follow-up
Many thanks to everyone who joined today's webinar Resilient Supply Chain Network Design presented by
We didn't have enough time to go through all submitted questions live during the webinar. In this article you will find our answers to the questions asked. Please be invited to keep the conversation going, right here on the AIMMS Community.
| Question | Answer(s) |
| How much time would it take to build a model and application like the one shown in the demo? I would expect it to take 25-30 days of building time for a not so experienced (AIMMS) model builder? | Models can be constructed quite quickly using an Excel data template and easy to use browser interface. The App is built and maintained by AIMMS, so you only need to worry about feeding it with meaningful data and running scenarios. Data should also not be a barrier : use what you have got, make some assumptions, and then refine from there. So within one week you should be looking at maps of your supply chain and running high level scenarios. |
| How to feed the model? Road network, location costs? | The transport and location costs are input data for the model. Data is ingested through an Excel template, and imported into the MySQL database using our Data Navigator App. |
| Is there any programming involved in the design process? | No programming is required. The App is built and maintained by AIMMS, so you only need to worry about feeding it with meaningful data and running scenarios. |
| What do you foresee as biggest challenge in current situation for medical device and healthcare supply planners and how should be their approach on inventory and forecasting postcovid ? | Sharp demand spikes for certain medical products (COVID related) and declines for other medical products (non COVID related). Artificial spikes due to panic building of inventory in warehouses and in homes. Supply disruptions due to lockdown closures of plants and warehouses in certain geographies. There is a lot of uncertainty around how things will recover and how quickly, so best case/worst case/expected case forecast and scenario modelling on these different possible view of the future is advisable. Don't depend on a single forecast number because it won't be correct. |
| Is this application available for pay per use basis? | Please contact Bhomik Jain for commercial information |
| Does the AIMMS can calculate the size of warehouse? | The app calculates throughput requirements for each warehouse. This can then be turned into a racking / pallet space requirement outside of the model using warehouse design expertise. Different warehouses have different conversions from throughput to physical size depending on type of product, type of racking, number of docks, etc., so it is difficult to provide any generic conversion. Often a % increase or decrease in throughput is already valuable information. |
| In our company, we do not change frequently.. hence will you do it for us? | One of our implementation partners can take on once off project type work. Please contact Bhomik Jain for more information in this regard. |
| Have you looked at a way to quantify resilience so that the model can be optimized on resilience instead of cost? | We see running multiple scenarios and preparing contingency plans as the best way to become resilient. The cost and service level KPI's of these scenarios is still important, but the real measure of resilience will be tested when disruption strikes, and could be meaningfully measured after the disruption has passed. |
| how do we know the cost to service from different locations to customers.. unless we do survey before hand? how do we address non availability of trucks? | The cost to serve report details the cost to serve for the customer in that scenario, and different scenarios can be run to see the change in cost to serve the same customer from different locations. Non availability of trucks can be modelled using maximum constraints on transport lanes. |
| Does AIMMS can only calculate the straight-line distance between different facilities based on the latitude and longitude? Because the actual truck travel distance maybe longer than the straight-line distance. | The app can use straight line distance (and duration based on average speed) OR actual drive distances and durations. A suggested methodology is to use straight lines for the initial exploration of scenarios, and actual distances on final scenarios. |
| How do you do the modelling for such network model? Is it on the web platform? | Initial data is ingested through an Excel template, and imported into the MySQL database using our Data Navigator app. Scenario runs are done from the browser, and you can also interact with the data through the browser. |
| How do we enable business users away from access to optimization tool to run their scenarios without depending on experts? | The easy to use Excel data template and intuitive browser screens are designed to facilitate easy use of the powerful optimization technology, so business users can run their own scenarios. |
| You assume reliable demand forecast data, but in reality there are significant uncertainties. How to deal with that in network design? | Model and validate a base case using historical data. Then model different demand scenarios from the browser or using different views of demand. To deal with uncertainty, model a worst case, best case and expected case, and ensure your network can manage all 3. |
| How fast will the model could handle a complex networks? | Runtimes vary depending on complexity of the data, and typically range between a few seconds to 15 minutes. For longer runtimes we help clients find more efficient ways to solve their problems. |
Contact details
Bhomik Jain
Bhomik.Jain@aimms.com
+65 91590659
Webinar recording
A recording of the session will be made available to everyone who registered. If you did not register for this webinar and would like to receive access to the recording as well please send an email to marketing@aimms.com.