Implementing new IT systems can be a risky business. But
Page 30
Page 31
Page 32
Page 33
If you've noticed an error in this article please click here to report it so we can fix it.
it doesn’t have to be like that, even for complex systems like vehicle routeing and scheduling software. Here’s how to do it
The chief executive probably reckons he makes the important decisions in a company. In fact, the really important ones probably come from the trafic ofice and from the load planners in particular. It is the quality of their work that drives transport eficiency. Unless the daily planning task is unusually simple, it surely makes sense to add technology to optimise the output.
Bearing in mind the well-documented dificulties of some big IT projects, we set out to ind the low risk way of selecting and implementing a daily routeing and scheduling system. This is not tracking, navigation or route optimisation. This is taking the company’s order book or workload and combining orders into sensible loads and routes; then sequencing deliveries to take account of criteria including cost, time, distance, delivery bookings/ restrictions, driver’s hours and access constraints.
We’ve taken advice from three well-known UK suppliers of truck routeing and scheduling software: Paragon Software Systems, Optrak Distribution Software and Route Monkey. Between them they have been involved with well over a thousand system implementations, so nobody knows better than them how to do it as painlessly as possible.
DO I NEED A COMPUTERISED SYSTEM?
At what point should truck operators consider moving from manual planning to computerised vehicle routeing and scheduling (CVRS)? Defining that moment in terms of fleet size might not be appropriate, because complexity of planning is the real deciding factor. But Optrak co-founder and MD Tim Pigden reckons CVRS typically becomes viable when fleet size is at least eight or nine and there are 100 or so orders to plan each day. He says the better definition is “when the projected savings are comfortably more than the cost of the system”.
Paragon MD William Salter also suggests that around 10 vehicles is often a watershed for introduction of CVRS. At Route Monkey, chief executive Colin Ferguson pitches the figure lower, reckoning that at about five vehicles “the savings in time, mileage and vehicle utilisation start to make sense versus the level of investment required”.
All agree that fleet size is not the only metric for considering a move from manual planning. Pigden mentions, for example, that planning failures such as missed delivery-time windows might be a trigger. “It could be simply that operators want to avoid too much reliance on individuals,” suggests Salter.
COMMON PITFALLS
Each of our three experts remarked on the fact that many operators embark on a CVRS project without a clear idea of their objectives. “They’re often rather vague,” says Salter, pointing out that this makes it hard for the software supplier to deliver the best solution. “We have to be prepared to do a lot of hand-holding.” Route Monkey’s Ferguson says: “Working with clients to understand a combination of parameters they may not even know themselves is the most challenging aspect.” It is not usually technical problems that threaten to make CVRS projects fraught: our experts say human factors are more likely to cause difficulties. Paragon’s Salter emphasises the need to “win hearts and minds” throughout the business, noting that a lack of commitment from senior management can lead to a troublesome implementation. “There is often resistance to change,” says Pigden, “particularly among drivers and planners. Drivers can be deeply suspicious of management.” Ferguson says: “Planners may see their jobs threatened. We pitch the system to them as an assist, not something that replaces them. It’s always going to need someone to manipulate the settings and add commonsense decisions.” The way to overcome people issues is to “get everyone on board,” says Optrak’s Pigden, adding that customers too should be told that their future deliveries might not follow the usual pattern. This is particularly true if your manual planning produces fixed or quasi-fixed routes; CVRS breaks the mould by examining more options. Make sure IT staff also buy into the project, “but don’t let the IT department’s agenda take over”, warns Salter.
WHAT ARE THE BENEFITS?
COST SAVINGS OF 20% CAN BE ACHIEVED, SAY EXPERTS Our three experts agree that CVRS systems typically deliver transport cost savings of 10%-20%. “Avoid unrealistic expectations,” warns Salter. “No software is a panacea. It can plan good, practical routes every day, but it won’t make the tea. Expect planners to make tweaks.” Pigden reckons there are instances where CVRS systems can achieve savings of more than 20%. “The disciplines of using CVRS mean that distribution data becomes much more systematic and visible,” he explains, suggesting this might prompt fundamental changes in distribution strategy. Typical benefits include:
● Fewer kilometres ● Better vehicle fill ● Less fuel used ● Greater driver productivity ● More on-time deliveries and better customer service ● Faster turnround times at delivery points ● Less planning time ● Later planning cut-off, reducing order lead times ● Better vehicle utilisation ● More consistent planning ● Easier to accommodate business growth ● A clearer overview of the distribution function
The advice is to use some key performance indicators (KPI) such as km/delivery to evaluate your manual planning. Agree realistic performance targets with the CVRS supplier. If all goes to plan, the capital payback period should be impressively short. “A few months,” says Pigden. “Within a year,” says Salter.
Route Monkey describes its system as “self-funding”, because the capital cost is usually spread over 36 months. “The monthly cost of our software is always less than the savings it generates,” says Ferguson.
CHOOSING THE RIGHT SUPPLIER
THE IMPORTANCE OF A TRIAL – AND A BACKGROUND CHECK Nobody should buy CVRS software without running a trial first, using actual data to compare one system with another and with manual planning. The size of the trial is open to conjecture: Paragon suggests it should be large enough to encompass a typical distribution cycle – a week, perhaps. Optrak suggests that one or two days’ data is normally sufficient to prove the case. Using more data adds to both parties’ workload without generating much additional useful information. Neither company charges for running the trial. Route Monkey says it sometimes carries out what it terms a ‘live trial’, running in parallel with manual planning. There would normally be a charge for this to cover Route Monkey’s time. The charge is deducted from the cost if the operator buys the software.
When scrutinising a trial’s outcome, operators should ensure that realistic settings have been used – results can be massaged by optimistic running speeds or rapid unloading times. Salter says trials also provide an opportunity to assess how easy it is to work with software companies. “A good supplier asks good questions and you can judge how responsive they are.” Buyers are advised to run background checks on software suppliers. This long-term relationship goes to the core of your business. Our experts suggest several ‘due diligence’ checks operators should carry out. They include: ● Inspect the supplier’s accounts to ensure it is a secure, viable business; ● Check references and visit existing customers whose planning task most closely resembles your own; ● Establish the quality of support: how many support staff are there?
● Check the supplier’s development resource and future development plans.
THE ABC OF ALGORITHMS
Sophisticated algorithms are the heart of CVRS systems, resolving constraints such as weight, cube, product compatibility and access restrictions to optimise loads and routes. But algorithms are deep within the system and our three suppliers said there are likely to be some significant differences between them. “The quality of the algorithms dictates the outcome,” says Ferguson.
Salter advises operators to test algorithms by getting into the nitty-gritty of the trial results. “A solution that on the surface seems to offer significant benefits is no good if the actual routes are unworkable on a daily basis, or if there is a genuine business constraint that can’t be handled,” he says.
Pigden homes in on the load-building algorithms, declaring these are where some of the biggest differences might be found. He reasons that achieving the very best vehicle fill is a fundamental starting point and believes that assessing fill by means of averages and approximations produces poorer results than calculating it by summing line-by-line three-dimensional measurements and weights. “If you’ve got lots of quite small things, you can probably get quite a good approximation with averaging, but that begins to break down with bigger and more awkward things.”
THE TECHY STUFF
Orders are input into the CVRS system by way of an interface with upstream systems such as sales order processing (SOP) or enterprise resource planning (ERP) systems. That should not be a problem according to our experts: all said they have interfaces to suit the popular order-handling systems.
Downstream connectivity is mainly to tracking and telematics systems, and this should be easy too. Paragon, for example, says it can interface with over 20 different tracking systems.
It is possible to use tracking to feed data back to the CVRS, thus comparing planned with actual, even perhaps making adjustments as the day unwinds – in theory. We are told there currently is limited demand for this, largely because there is little opportunity to change things. However, Salter mentions a big retailer with a 24-hours-a-day operation that gradually slips from plan if minor delays begin to accumulate. It uses real-time tracking data to ‘re-spin the plan’ to get actual and plan synchronised once more.
We all know software development moves on apace. So, is a CVRS system heading for obsolescence within a few years? Should operators attach a finite life to planning systems? Our suppliers say annual subscription charges cover the cost of software upgrades and new map releases, so systems should be good for many years.
IMPLEMENTATION
“Implementation is all about planning,” says Salter. “Operators should take advice from their software supplier – we’ve done it hundreds of times before. We have a tried-andtested methodology. We draw up a detailed implementation plan and make sure everyone knows what their role is.” It is typically two to four months between placing the order and CVRS going live. Suppliers reckon project progress is normally determined by the customer, in particular the time taken to ‘cleanse’ data such as product information, postcodes and delivery restrictions.
Once the software supplier has customised the CVRS system to your exact requirements – some are more flexible than others – the software is installed on your computer and tested to check the interfaces. Then it is time for staff training, followed by first use of the system. Optrak and Paragon advocate a phased introduction, ideally starting with a period of parallel running alongside the manual system to iron out any glitches. “A common problem is that clients find it hard to allocate sufficient time to the parallel testing,” says Pigden. “After all, the planners are still planning in the old way.” He believes management must be prepared to allocate the resource to deal with this, arguing that it makes no sense to delay a system expected to save money.
Ferguson says the original trial should have generated confidence in the system. “Our customers wouldn’t be going ahead with the system if the trial didn’t go well,” he reasons. He advises small companies to go for ‘big bang’ implementation, with phased roll-out more appropriate to bigger projects.
COSTS
WHAT DO YOU PAY – AND WHAT DO YOU GET FOR YOUR MONEY?
With such rapid payback periods, the cost of a CVRS system is almost immaterial. But in a world ruled by budgets, this issue must be addressed. A typical Paragon system for 20 vehicles in a single depot would cost around £20,000. That rises to about £50,000 for 100 vehicles spread across several depots. A multi-depot system for up to 1,000 vehicles would be around £250,000. These are licence fees for the software. Then there is the cost of training. Salter says a typical training package (for as many staff members as you wish) is normally 10 to 12 days, costing around £7,000.
Finally, there is the annual maintenance fee. Paragon’s is levied at 15% of the licence fee and covers technical support, map updates and software upgrades. Salter advises operators to check the quality of telephone support. “You don’t want a triage service where somebody rings back later,” he says. “You must get through to somebody who knows the system.” If you plan at night or weekends check availability and cost of out-of-hours support.
A typical “entry-level” Route Monkey system for fleets of about 25 vehicles costs around £10,000. This relatively low starting point makes Route Monkey attractive to smaller fleets. “That was the gap in the market we identified,” says Ferguson. The precise cost depends on development time, so a small fleet with complex planning could pay more than a larger fleet with simpler needs. Payments may be spread, typically over 36 months. Training is included in this capital cost. Route Monkey also charges an annual fee of 15% of the original capital cost for support and upgrades.
Optrak’s charges are structured differently. There is a one-off implementation charge of £10,000 to £15,000 for a system suitable for a typical 25-vehicle fleet, with an annual subscription charge, also £10,000 to £15,000, for support and upgrades. Pigden explains this relatively high subscription charge means that chasing new business never overshadows supporting existing clients.
There are several web-based CVRS products that provide online access to planning software. This eliminates the capital purchase, with all costs covered by a monthly service charge. Although the payment system is different, the selection and application principles are similar.
Top tips for selecting and implementing CVRS software
● Be clear about your objectives ● Give potential software suppliers accurate information and data ● Have realistic expectations ● Use KPIs to evaluate performance before and after implementation ● Use due diligence when selecting a supplier ● Scrutinise trial results to ensure they are robust and realistic ● Understand and manage resistance to change in your company ● Understand the full costs but focus on the operational benefits ● Take up supplier references with other appropriate operators ● Plan and manage implementation carefully – heed the software supplier’s advice ● Do not skimp on training ● Ensure there is good support when you need it ● Examine the supplier’s resources and commitment to ongoing development