I got into the habit of cycling to work after the pandemic. The cool morning breeze, the first rays of sunlight in my eyes, what a great Feeling! My efforts have been reduced since I switched to an electric bike. One morning, coming out of the cover of undergrowth, round a bend, still in darkness, I was driving at high speed. An unscrupulous squirrel suddenly crossed the bike path. Surprised, my reflexes prompted me to attempt a dodge which unfortunately took me to one side. I awkwardly regained my balance and resumed on my way.
You will ask me, “But what is the connection between swerving a bike and demand sensing and supply chain planning in general?” Here are some explanations I can offer you.
Demand sensing is the term used to describe the concept of short-term demand management. Defining the most important factors influencing the market, monitoring these indicators and incorporating them into more or less complex models are some of the elements that define demand sensing. We must not forget the various human factors that are difficult to quantify. Are these concepts new? No. Simply using the new vocabulary makes them specific, stimulates the discussion and introduces possibilities.
A Strategic Vision
When I first invested in my electric bike, I had in mind that I would be able to ride at weekends and, if possible, get to work. Of course this desire was not a whim or just wishful thinking. It was a well thought out projection for a means of transport that would have many benefits to me over the next few years.
Like any business, my long-term plan was outlined. Of course, I had to refine this need to make my purchase a reality. For what use, in what conditions, with what budget? During this strategic period, I may have considered some points of the home-office journey in terms of route and variants. Whilst there is no doubt about the fact that at seven o’clock in the morning with the sun coming out of the undergrowth we have a backlight which prevents good visibility, this detail was not considered.
When a business draws up its strategic plan, it doesn’t care whether it will be cloudy or sunny tomorrow. It is also not interested in whether the current promotion in a particular supermarket is going as planned or not. It asks itself other, broader, questions and all these questions aim to define a long-term plan.
Test Different Tactics
Once my investment was made, and my new means of transport received, I could consider testing different routes to get to work. My first interest was to favor cycle paths over roads congested with vehicles of all kinds. So I was able to test 4 or 5 roads to decide on the one that I will favor. Of course this is the road through this famous undergrowth.
For any business, the routes to achieve a long-term plan are not fixed in advance. This business will check many variations before establishing its path. In any execution of a plan, the company may question these tactical choices and must adjust to market information. Establishing a medium-term forecast allows the company to go beyond the simple figure generated. It can ask questions about its environment and its evolution. How to position itself and react to changing situations. It’s just a matter of making assumptions and testing them.
React and Adapt
That morning, after checking my electric bike – Battery charged, tires inflated and brakes operational, I took the road to work serenely. it was an unfortunate coincidence – the backlight, the brisk pace, the squirrel crossing the bike path – that caused me to swerve. Ah! if only I had the information I needed to prevent this. I don’t know, a sensor could have informed me of the presence of an animal, another could have controlled my swerve to avoid the low side and finally another could have acted on my brakes to slow my speed.
For any business, short-term events affect the implementation of its plans. Some can be dramatic. Without wishing to prejudge having all the information necessary to run a business, I am firmly convinced that certain basic ideas can guide you:
- It is important to identify the external elements that most influence your business.
- It is important to put in place the right “sensors” to identify and measure these elements.
- You must build the models of sensible reaction, adapt your tactics and keep the strategic plan in sight.
Here I am describing the basic recipe for “demand sensing”. Of course, the more event “sensors” there are, the more complete and complex your model is. Manual piloting becomes less and less possible. For this, the mathematical tools of artificial intelligence allows multiple elements to be taken into account together to give you sensible indications.
“Sensors” and the information that derives from them, must be constantly challenged. In the same way that a wastewater treatment plant is self-controlled by turbidity, oxidation reduction, video recognition and olfactory sensing, it is not uncommon to have to reset these sensors when they start to drift. This is when we need a person to intervene. Also, in the case when a squirrel appears…
Solutions exist. For example tools developed by QAD DynaSys are able to take into account information from “sensors”, interpret them and provide sensible reactions.
You will understand that many questions are opened by these simple reflections: could we really do without people in business decision support? How much confidence should we have in artificial intelligence models? A self-driven supply chain… Is it a dream or a possible reality?