Selecting a Computerised Vehicle Routing and Scheduling (CVRS) System – 101 | Plan Forge

Why did people start to focus in this area? Transport costs were increasing and service levels were becoming a focus. Companies needed to get their distribution department in order. The best way to cut costs? The good old term “sweat the assets” and I will add “you need”! The best way to improve service levels? Have a plan that is accurate and can be fed back to the customers. Over all increase utilisation through proper planning.

Today in South Africa there are over 6 major contenders for Computerised Vehicle Routing and Scheduling systems in the market. So which tool is best? What does each one do? Where do you start to answer these questions when looking for a system.

The place to start, is to understand your own needs and requirements before even going to market to enquire about software.

The first question you should ask is: Do I need a Vehicle Routing and Scheduling system?

Most of these systems are based on optimisation algorithms. So the smaller the problem i.e. number of orders and vehiles, the less there is to optimise. So if you have 1-5 vehicles the potential optimisation is very little, and more than likely, unless you have vehicles standing around, any savings will be achieved from better control of the planning process through to execution of the plan. So if you want greater control, better visibility and are looking to expand, then the answer is yes. The alternative is to increase the control through other means and the answer would be no.

Next you should ask: What problem are you trying to solve by getting a CVRS?

This is an important Question. It will lead to many different answers both on a strategic level and a functionality level.
If the strategic answer is: Cost savings and better control from planning to execution. Then you are on the right track. However if it is to give drivers directions or help with geographic locations of customers etc. then there are other lower end systems, which can help you with this, that contain no optimisation algorithms.

The main objective of a CVRS system is to save cost through improved utilisation using an optimisation engine.

This comes in the form of reduction in the number of vehicles used, better time utilisation, better capacity utilisation, and less mileage and time per unit delivered.There will be more detailed functionality that you require from which ever CVRS you select. This could be from multiple-depot planning capability to static route planning. Below are some of the major pieces of functionality to be aware of.

Primary or Secondary Planning: This is the first major split in the decision making when looking at a CVRS. Some people debate that there should be no distinction between the two, however in the SA environment there are some clear differences in characteristics between them. Plus there are systems in the market, which strengths are clearly in either Primary Planning or Secondary Planning.

Primary loads are usually long haul loads with usually no more than 3 drops on the vehicle with full pallets, either going to other depots or large volume delivery points. Plus this type of distribution is usually out-sourced to hauliers. The rates charged for these deliveries are also structured on a “rate by lane” basis with additional extra drop charges and demurrage. Date horizon, rate selection and backhaul optimisation play key parts to primary planning. It is also usually a central function.

Secondary loads on the other hand are usually 24 or 48 hour lead time local area deliveries within 100km. They are multi-pick multi-drop loads focusing on the store delivery. Time and vehicle utilisation play a key part in this optimisation. And all orders received on a day, must usually be planned for delivery within 48 hours. The key cost drivers for secondary planning are a cost per hour and a cost per kilometre therefore increasing the utilisation of the vehicles. Secondary planning is usually a local transport planning function.

There are however some grey areas where the two can overlap.

Lets focus on primary planning for the moment and the main functionality that is required.

• Multiple client infrastructure: To achieve optimisation with a primary planning tool you would want to include as many loads as possible. As optimisation opportunity is limited to the number of orders you will be planning and those orders origins and destinations. Therefore the more clients you can include out side of your current business the more optimisation opportunities there are. This is usually a non-competitive part of any business and cost is a large driver therefore it will outweigh any competitive obstacles. Many bureau services run primary planning tools.

• Multiple Hauiler and Rate Capacity: One of the key benefits to have a primary planning tool is for the tool to hold all the possible rates that can be attained on any lane, thereby empowering the planner to make a quick and correct decision based on costs and service.

• Central planning platform: As primary planning will more than likely be a central function to achieve the best visibility. Whichever tool is selected must have a client/server infrastructure to allow for multi-user functionality. This will give the visibility across the organisation of the primary planning activities.

• Optimisation Engine: This is really an optional depending on the volume of orders planned on a daily basis. You can achieve 80% of the benefit of a central planning tool by using it to plan manually. This is due to the benefits of primary planning mentioned above. However when the sheer volume of planning gets too much an optimisation engine can give the automated load building that can save time and add the next 10-20% of benefit that can be achieved.

• In-house or Outsource: This is not really about functionality but rather costs decisions. Primary planning tools are generally very expensive to purchase and implement. Therefore if the volume of orders that you need to plan does not warrant the costs, there are bureau services that currently do primary planning for multiple customers. The benefit of this is that there will be no capital outlay plus you would be achieving better utilisation immediately by being added to a larger primary planning community.

For secondary planning the requirements are vastly different. There are also 2 types of secondary planning. There is dynamic planning, which is a different set of delivery points every day which need to planned. Then there is the static routing scenario, where the same delivery points need to be serviced everyday. The proverbial “milk-run”.

• Speed of planning: Due to the time pressure that is usually in the dynamic secondary planning environment, picking is usually reliant on the transport plan. The transport plan is best if a single pool of orders is optimised and this can only take place once order capture has been completed and cut off for the day. So the planning is between the proverbial “rock and hard place”. Due to this an optimisation engine is required. This speeds up the planning and if the system is correctly configured very little amendments will need to be made therefore keeping the planning time down to a minimum.

• Optimisation Engine: Over and above the speed of planning aspect for dynamic routing, the optimisation engine is critical to all secondary planning. There are two reasons for this. Firstly the optimisation engines can be customised with different parameters to reflect how you want them to optimise based on costs and time. This also gives you more control over your secondary planning. And secondly it is required to get the best possible answer in a short space of time. For example the number of possible permutations for just 5 drop points is 120! If you are planning between 20 and 200 orders per day it is physically impossible for any human to get the best possible plan mathematically based on time and cost even given one day to do it, let alone one hour. An optimisation engine can.

Advanced planning options: This depends on the requirements on the business, however listed below are some of the more common requirements and features.

• Zone or territory control: In certain planning problems it may be necessary to break the problem down, either geographically or otherwise i.e. distribution channels for supermarkets vs. corner stores. A territory or zoning function will force delivery points with the attached zone or territory to be planned on different loads. This is a very useful piece of functionality in any planning system.

• Delivery point restrictions: This is a common requirement. Restrictions should include vehicle type restrictions as well as delivery day and time requirements.

• What-if Functionality: Check to see if the system allows you to use multiple datasets. This is key for static secondary planning where you will want to run several scenarios for comparison. I.e. considering night deliveries. There are many scenarios where you can then use your operational data to help with strategic decision-making.

• Service Day Assignment or Territory Balancing: Some secondary tools have more tactical capabilities, which may be of interest when doing the investigations. Enquire about any tactical or strategic planning modules they may have available.

• Actual feedback reporting: One of the most important parts of a secondary planning system is keeping it up to date or calibrated. I.e. taking the actual completed trip sheet and comparing it to the plan. Any variances should be analysed and catered for. This will allow the planning to be more accurate in the future.

• Once you have decided that you require a CVRS and have detailed down your requirements for the functionality it will be time to go to the vendors and let them take you through the various products that are available. However there is one more question that you will need to ask.

Am I serious about using a CVRS?

Once you have made the decision to invest in a CVRS, this is the most important question. I have done many implementations of CVRS systems over the last 8 years and the ones, which always succeed, are the ones which the client dedicates proper resources and takes full ownership of the project. At the end of the day you have purchased the system and own it, not the vendor. A CVRS is not a magic box that will solve all your problems. As with any other IT system it requires information to work and processes to deliver results. If the correct focus is put into those two areas, success is imminent. If not failure is inevitable.

To all those of you who have a CVRS but don’t use it, consider dusting it off and giving it another try. The reasons the implementation failed before may not be applicable anymore. Plus to achieve 80% benefit even if you want 100% is better than nothing!

By Andrew Stephens from NeoLogic Solutions