Expect artificial intelligence to impact all aspects of the fulfillment process.
“If it ain’t late, you ain’t making money.”
Sound familiar? This has been a decades-long mantra of ready mixed concrete (RMC) dispatchers all over the world when pressured to minimize the typical 17 to 23 percent of total costs due to logistics. And the bookend customer complaint to dispatchers has universally been, “Don’t lie to me about lying to me!”
Through the advent of mobile technology combined with powerful artificial intelligence (AI) programming, the customer can now see everything spanning truth and lies. Woe be upon the dispatcher who continues to embellish delivery and truck statuses—yet the RMC producer still needs to make money.
The bigger picture
Mathematical optimization for RMC logistics has a history of mixed results, but the vast potential for very real reward persists. Let’s foreshadow the topic with help from the late W. Edwards Deming, an electrical engineer and mathematical physicist of world-renowned talent.
Deming famously stated, “Sub-optimization is when everyone is for himself. Optimization is when everyone is working to help the company.” In our world, if we optimize logistics for a truck (the individual), we improve the truck’s performance. If we optimize logistics for the collection of orders (the company), we improve the company’s performance.
Caution! Paradigm shift required
Manual dispatch for RMC began decades ago. Back then, old-timers—often gruff and chain-smoking and sometimes prematurely aged due to the pressure—worked with a large board on the wall representing the plant, with rows (trucks) and columns of hooks (time slots) to hold index cards representing the live logistics. While the advent of computers initially expanded the capacity and speed of the dispatch process, the programs merely automated paradigms of the manual method. The board was simply moved to a computer screen. Real-time delivery ordering and truck-tracking was still done by phone.
Faster computers with more memory eventually allowed additional factors to be considered and information to be managed electronically, but they still did not substantially change the truck-scheduling process. Enter “computerization of optimization,” a collection of mathematical principles and methods for solving quantitative problems. Faster computers combined with modern AI substantially increase the reach of enhanced processes, particularly for the complicated and large challenges in RMC.
Alas, once again, humans are the stumbling block. We insisted on truck optimization, which diminished the impact on company performance. Subsequently, there is an industrywide acknowledgment that many prior attempts at RMC optimization fell short of expectations. We must elevate our thinking to the collection of orders.
Paradigm shift, applied
INFORM Gmbh, based in Aachen, Germany, has developed an AI optimization package called Syncrotess. They have proven platforms dedicated to such things as automotive manufacturing, airline logistics, fraud detection, and—you guessed it—RMC logistics. With subscribers worldwide, they have extensive use-case evidence of the financial impact of order-based optimization.
While INFORM works with the biggest RMC producers, let’s start small and consider a representative, real-life proof point of an EU-based, 37-truck operation. The INFORM customer is augmented with dedicated independent and wildcat independent trucks.
For the target day, the company’s manual truck-based dispatch results in USD were:
- 37 owned trucks with fixed costs that never sleep: $13,126.
- 23 owned trucks deployed had a variable cost of $3,936.
- 20 dedicated independent trucks: $9,705.
- 10 wildcat independent trucks: $8,700.
- Total manual dispatch logistics trucking costs: $35,467.
The INFORM optimized process with all owned trucks deployed resulted in:
- All 37 owned trucks deployed: fixed costs $13,126.
- Owned trucks’ variable costs: $4,868.
- 19 dedicated independent trucks: $11,921.
- Total optimized logistics trucking costs: $29,915.
The result was a $5,552 (15.7 percent) reduction of transport costs for the day. Of course, part of the cost reduction was putting the schedule in place sufficiently before the prior day’s close to call up all company-owned trucks. However, the bottom line still counts.
Nibbles, not bytes
AI will impact all aspects of the RMC fulfillment process. The most obvious will be a constant stream of small suggestions—nibbles, if you will—to the customer, dispatcher, batcher, etc. For example, the platform’s message may state: “Hey finisher, you are placing an order for outside flatwork with a 15-minute spacing. Your last six orders requested 15 minutes, but you actually performed at 25-minute spacing. How about changing this time?”
Dispatch also will be impacted, but in the isolated paradigm of the action at hand. For instance: “Hey dispatcher, have you considered splitting the order 50/50 between plants 4 and 7? This will reduce unused truck time and deadheading by 73 minutes.” While much better than straight manual dispatch, it will, by definition, be much less impactful for cost savings than a full-scale order optimization model.
The “Hey (sales guy, finisher, batcher, etc.), have you …” model will become the default baseline for all aspects of a producer, with the caveat that dispatch is much better served by the big, more complex order approach. Consider the comparison between a jet airliner and a helicopter. The jet will always be faster and more fuel-efficient, but the helicopter will be the only practical, albeit much more limited option for short-distance flexibility. (Full disclosure: Your columnist leads a company that provides AI for exactly this purpose.)
ON-TIME DELIVERY VERSUS TRUCK PRODUCTIVITY
Is this trip worth it? Comparison to cement optimization
Consider a mix with 400 pounds of ordinary portland cement (OPC) at $150 per short ton (ST). Let’s say the cost of the supplementary cementitious materials to replace the OPC is $30/ST. For every 1-percent reduction of OPC through optimization, the net monetary savings per cubic yard (CY) will be ~$0.24. Given 8.25 CY per load, the savings would be ~$2 per load.
Transport is typically between 17 to 23 percent of the total cost structure of RMC. For a total cost of $125 per CY and using the low end of 17 percent as the transport cost, that is $21.25. If we can save the low end of the exhaustive INFORM optimization proof points of 15 percent, that’s $3.19/CY—and at 8.25 CY per load, that’s $26.30 per load. The notion of low-hanging fruit comes to mind, as it will be difficult to find a better return on working capital.
Looking forward
My dentist has a framed picture in the waiting room with the caption, “Only floss the teeth you want to keep.” Likewise, Deming wisely proclaimed that the “two basic rules of life are (1) change is inevitable and (2) everybody resists change,” followed by another wry observation: “It is not necessary to change. Survival is not mandatory.”
Our beloved industry is in the same boat. RMC producers do not have to change. However, those who survive will.
Craig Yeack has held leadership positions with both construction materials producers and software providers. He is co-founder of BCMI Corp. (the Bulk Construction Materials Initiative), which is dedicated to reinventing the construction materials business with modern mobile and cloud-based tools. His Tech Talk column—named best column by the Construction Media Alliance in 2018—focuses on concise, actionable ideas to improve financial performance for ready-mix producers. He can be reached at [email protected].