Implementing an AGV System
Types of AGV Companies
AGV Vendor Analysis
How to Approach a Potential AGV Application
Should You Work with a Consultant or an AGV Vendor?
Obtaining AGV System Proposals
Product, Layout and Throughput Rates
Degree of Automation
Comparisons in Financial Strength and Technology
Who Should Go?
Questions: What to Observe
There are several important steps that need to be followed when implementing a system. Use the Investigating an AGV System to help you determine your viability as a user. Use the Guidelines for Successful Systems to help you determine common pitfalls, problems and solutions.
Guidelines for Successful Systems
AGV systems are being used in a growing variety of applications than ever before. Where they were once used only for warehousing and long distance transportation, today they are now being used as short move transfer cars, in-process transportation and as an integrated part of on-line manufacturing processes.
Nowadays as you evaluate a potential material transport application, you will not only consider conveyor, mechanical transfer cars, or conventional forklift, but also AGVS. Selecting AGVS over other solutions is only the first step. You will need to spend a lot of time deciding the scope and functionality of your AGV system as well as which AGV company is best suited to be your supplier.
Types of AGV Companies
AGV companies come is a variety of types and sizes. Today the market has developed beyond just general application types and into specific niches. One company does not cover all niches equally well. Therefore, it is wise to evaluate AGV companies based on a number of factors. Those factors should range from not only what they currently offer, but also what has been their history and do they have the resources to provide you with what you specifically require and have the ability to support your system in the long-term.
These are AGV system suppliers that are generally small companies with less than $10M in sales and 15-50 employees. Typically, these companies operate in a niche market such as textiles, paper, light assembly, custom design, etc. They tend to offer simpler type AGVs. Most single product AGV companies confine their sales only to the US market. A single product type AGV company will usually be focused on providing AGV systems that operate as an stand-alone system. That is, the system is not integrated with other automated systems under a common controls platform/network. Stand-alone systems provided dedicated AGV service without directly linking to other automated systems. Typically, these systems require less complex controls because they provide dedicated and self-contained functionality.
Multi-Product – Multi-product AGV companies are those that offer products other than just AGVs. These companies are generally much larger than a single product company and can have total product sales ranging from 50 million to many billions of dollars. Companies in this category operate in the domestic market as well as internationally. Multi-product companies have a further characterization. They may provide either ‘stand-alone’ AGV systems and/or ‘integrated’ AGV systems.
Stand-alone System – Even though a multi-product company is typically a large corporation, that does not mean that they will undertake any type of AGV project. Some will more or less exclusively concentrate on ‘stand-alone’ AGV type systems. Those systems would either directly interface with people or with conveyor load transfer locations.
In addition to supplying stand-alone systems, some companies in the multi-product category also provide ‘integrated’ systems. AGVs operating in an integrated system environment typically interface with automated storage and retrieval systems (ASRS), conveyor systems or other automated systems. In these cases, the AGV system controller is networked as part of an overall, integrated material handling system. Product can therefore be tracked throughout the integrated system and since there is a common control system, it is easier to optimize the total system performance.
AGV Vendor Analysis
Below is a list of 25 questions designed to help you compare the different technologies of AGV vendors.
1.How does your wireless technology work and when was it introduced?
2. How is AGVS vehicle routing accomplished in your proposed system? What controls are required in the floor, on the walls or columns, or on the vehicle to accomplish vehicle routing?
3. How is vehicle traffic control accomplished in your system? What controls are required in the floor, on the walls or columns, or on the vehicle to accomplish vehicle traffic control?
4. Does your system employ a central AGVS controller?
5. Do you have a system monitoring CRT graphics display available? Is it included in the proposed system price? How does the terminal operate? What are the monitoring, recording and command issuing capabilities?
6. What are the operational steps to remove or enter a vehicle in the system? What are the rules a manual operator must follow to safely remove or enter a vehicle on the system? Where can this be done? What controls need to be reset or cleared?
7. What hardware controls are part of the floor path network? What is the quantity of guidepath wire energizers, traffic controllers, routing controllers and floor code markers (wire, plate, magnetic or responder type)?
8. Describe in detail manual operation of the vehicle off the path. How does the operator drive the vehicle manually?
9. Can a vehicle be entered or removed from the guidepath at any location, including the middle of an intersection, without resetting traffic controls?
10. Is the system as proposed? Is it configured with zone blocking traffic control throughout the guidepath network including all straight-aways?
11. What other additional types of vehicles can be used on the same initial base guidepath with the base vehicles? Are there any additional system changes which may be necessary to accommodate different types of vehicles?
12. What are the necessary steps to expand the system with additional vehicles and/or guidepath? What possible additional controls, rewiring or software are required to add vehicles, stops or guidepath to the base system?
13. How can the proposed system be upgraded to higher-level computer controls? If a central system management control is employed in the future to automatically dispatch vehicles, what additional hardware or software is required of your system to interface to this higher level of control?
14. Explain communication capability to and from the vehicles for remote vehicle dispatching. Is this employed in the base level operation or is it a capability which needs to be added to the system when a higher-level computer management system is added?
15. If you offer gyro-based wireless AGV navigation, explain what type of markers are used in the floor and typical spacing.
16. Describe the user interface with the onboard AGV control panel. What functions are provided?
17. How many AGV systems have you installed?
18. How many years have you marketed an AGV product?
19. Is the product of your own manufacture? Are any parts licensed or procured from other sources? Please explain.
20. If offering a unit load carrier type AGV, explain the construction, controls and steering mechanics of the vehicle.
21. Where are your AGVs manufactured?
22. Where was AGVS technology developed? Where is the corporate AGVS product group located?
23. Where will parts and service originate?
24. Where does documentation and training support come from?
25. How long have you offered AGVs in the US and how many AGV systems do you have installed in the US?
How to Approach a Potential AGV Application
Investigating AGVS for a possible application need not be a costly or time-consuming task. All AGV companies have ‘questionnaires’ or ‘data forms’ that ask for basic information regarding the potential application. An initial system concept and budget price can be generated once this short document is completed and sent to an AGV company.
Don’t limit yourself to a single vendor at this stage. Get feedback from several to insure you get a fair picture of what AGV possibilities exist for your application. Some AGV companies may prefer a given type of AGV (tow, unit load carrier or fork type). This may be because they only offer one or two types. Others with broader product lines may offer a different type of AGV for the same application.
There are also cost and benefit tradeoffs to consider. A given type of AGV may result in a more expensive project cost, but may provide a better payback because it saves more money in the long run. However, don’t always assume that the more capable AGV system is the better system for your application. Capability comes at a price other than dollars. Often a more capable AGV is a more complicated AGV and this could adversely affect maintenance costs and reliability. It may be better to do a little less with the AGV system when real world issues (user friendliness, plant maintenance ability, problem recovery, environment, etc.) are considered.
Should You Work with a Consultant or an AGV Vendor?
The chances are that if you are planning a complex or integrated type of AGV system, you will employ a consultant to assist you with the planning. However, in general that is not necessary for most other AGVS projects.
Among the factors, beyond an integrated or complex system that need to be considered concerning use of a consultant or not are:
- Data gathering – do you have the time to do this?
- Cost/benefit analysis – can you prepare one?
- System concept evaluation – do you have resources to compare pros/cons of different AGV approaches?
- Vendor analysis – are you capable of evaluating different AGV vendors (experience, reputation, resources, etc.)?
It is probable that over 90% of the AGV systems installed each year do not involve direct consultant assistance. Most AGV systems have a clearly defined mission and are generally the stand-alone type. This being the case, most customers prefer to approach the use of an AGV system with internal planning and evaluation resources.
As a rule, good common sense is the best resource you can have. And, you usually don’t need to pay for it if you already possess it. Users will find that AGV companies have the genuine product application experience that most consultants lack. Appropriately blending good old common sense with product application experience is proven recipe for a successful AGV system.
Obtaining AGV System Proposals
Depending on your circumstances, you may approach obtaining vendor proposals in different ways. The proposal stage of an AGV project is an ideal time to interact with an AGV vendor. This experience is an often overlooked opportunity to ‘size up’ a vendor with respect to intangibles such as quality relationship, communication, integrity and support. These values may mean the difference between a rewarding experience and an unnecessarily challenging one.
Method #1 – Best Concept Approach
The most common practice for obtaining proposals for an AGV project is to permit each invited vendor to recommend the best concept. Remember that AGVs are not commodities. Therefore, each AGV company has different product and technology to work offer. What may be a strong point, may be a weak point for another. Letting each make their own recommendations opens up the possibilities to see alternate concepts which may be better than the initial concept.
To implement this approach, you will need to do more up front with each of your invited bidders. In many cases, it makes sense to screen the potential vendors after an initial concepting/budgeting phase. Then work more closely with the finalists to provide them with an adequate opportunity and information to develop their best concept to bid. Your main responsibility is to insure each bidder is provided with the same set of requirements without dictating a solution to them.
Method #2 – Specified Concept Approach
There are times when it is preferable to specify an AGV solution to a set of invited bidders. More often than not, this occurs when a consultant is involved in the process. Usually, the consultant develops the AGV system concept working with you and then translates it into a bid specification document.
The advantage of this approach is in working with an unbiased consultant to evaluate the risk/benefit rewards of various concepts before deciding on an approach to be specified in the bid request document. Each bidder can then be judged as to how well their proposal response meets the evaluation criteria established for the specified concept.
A consultant does not have to be involved in a ‘specified concept’ approach type AGV procurement process. Should a potential user be so inclined, they may develop their own bid specification. This document would define the system concept to be bid in such detail that all invited bidders would have to provide the necessary controls and equipment to meet the specific concept requirements. Typically, this restricts the bidders to proposing the same type of equipment and similar controls system capabilities.
The downside of bidding a specified concept is that unless there is an overriding reason to require a high level of bid conformity, the end user may well loose the opportunity to benefit from each bidders unique concepting expertise.
Developing the AGV system concept can be a short or a long process depending on the objectives for the project. Many AGV systems have a dedicated, single task. This makes detailing the concept very straightforward. Systems that include many separate tasks, generally require considerable concept development time. An advantage of AGV systems is that they do not have to be implemented with the ultimate scope included. You may start with a single task and easily expand both the scope to the task and add new tasks at any time. This is particularly important if you are concerned with the initial introduction of an AGV system into your facility.
Concept Validation – If a concept is inherently simple and straightforward, validation is usually pretty easy. However, the more you wish the AGV system concept to achieve, the more important that you validate it in several ways.
Benchmark Validation – The easiest validation method is to look for similar previously installed systems. Benchmark performance and operational factors of those systems to determine if your concept will achieve your expectations. You may find that you have over or under estimated some aspects of your concept. Adjustments can easily be made at this stage.
User Validation – Never underestimate the importance of user validation. The people on the floor where the AGV system will operate often are the best source for valuable feedback on a system concept. These people are familiar with day to day operations and are in the best position to tell you whether a concept has flaws and how to address them. Users are very good at identifying ergonomic, space, interface and operational issues.
The experience of the users can provide the foresight to avoid aspects of a concept that do not take into account real world issues. By having users ‘walk through’ the concept you will not only gain their insight but also their support for the system when it is installed.
Simulation Model Validation – Not all AGV system concepts require simulation validation. However, building a computer simulation model of the system is often valuable if only to graphically illustrate the system concept in operation.
Involved systems or systems required to meet strict throughput rates should always be simulated. The only way to insure this type of AGV concept will perform as expected is to perform a computer validation simulation. Building the model forces decisions on exactly how a system should be structured to operate. In affect this causes the planner to address these issues soon rather than later. Any problems with the results of the simulation are immediately addressed by changing these operational issues.
A further benefit of a computer simulation to validate a system concept is that it visually demonstrates the system concept in operation before it is installed. This can be a valuable aid in gaining general support for the concept.
One must remember that a computer simulation is only as good as the data it employs. Even when this is the case, there is no guarantee that the concept will perform as simulated unless all the necessary factors from the simulation are replicated in the actual system implementation.
When all is said and done, whether you chose to utilize ‘benchmark’, ‘user’, or ‘simulation’ validation, a sanity check should always be performed. This can be achieved in a variety of ways.
Plain old common sense is an often-overlooked virtue. Sometimes people get too wrapped up in automation to realize that they are actually making a concept too risky or complex.
Some simple general questions to ask are:
- What is the concept payback?
- How much support will the concept require from the supervisory management level.
- What are the maintenance requirements for the concept?
- Are the load rates required from each AGV excessive compared with real world experience?
- Are we automating an interface that provides no benefit?
- Can anything in the concept be simplified?
- Don’t overlook asking the ‘what if’ questions such as:
- What if our rates change, how will the system adjust?
- What if an operator fails to unload an AGV or AGV station?
- What if the central AGV controller fails, how do we operate until we recover?
- What if an interface fails, how do we handle it?
- What if an operator interferes with an AGV, how do we recover?
Product, Layout and Throughput Rates
The heart of any potential project is the identification of the requirements. Providing too much information can be as bad as providing too little. Most AGV companies are looking for basic parameters from which they can build a concept and develop a proposal. The basic information that needs to be provided includes:
Product Handled – What is the product? What are the dimensions and weight? How stable is the product? Is the product to be loaded manually on the AGV or by automated interface?
Identify all products to be handled. In addition, provide accurate footprint data for the load container (pallet, rack, etc.).
Layout – Provide a drawing (CAD layout, preferable) for the facility. Identify where the AGV stations are to be located. If you wish, create the AGV path linking the stations, or the AGV supplier will recommend a route in his proposal.
Throughput rates – It is very important to supply load transport rates between the various AGV stations. This can be accomplished by filling out a simple matrix such as the one provided below.
Typically these rates are specified in units per hour. You should specify the maximum rates and the normal rates. It is important to indicate how often the maximum rates occur, and for how long they are sustained. Also, do the maximum rates for each station occur all at the same time? You would not want to provide enough AGVs to handling a simultaneous maximum rate from all stations if probability of that occurring is very low. It is important to decide what sustainable rate you need to support and what maximum rate you want to handle for a limited time duration.
Degree of Automation
How much automation you want is often different from how much automation is practical in your environment. It is easy to over-automate when the benefit to risk ratio is turning negative.
As a real world matter, it is always necessary to consider how the system’s users will interact with the system. Too little automation causes the users to do too much and the system benefit may be marginalized. However, too much automation may cost the users too much in flexibility.
A common misconception is that ‘automation increases flexibility’. In truth, the opposite is often the case. A highly automated system imposes a very narrow operational methodology limiting user flexibility. This is good if the process supported is highly repetitive and regular as is the case where an AGV links two or more other automated processes. In the case where the AGV system is intended to interface directly with users and conditions of use may vary, it is a good idea to keep the level of AGV automation to a more reasonable level. Doing so will allow the users to accommodate varying conditions without being locked in a rigid and maybe less efficient operational regimen.
Controls/User Interface – This area is somewhat related to the degree of automation issue, but is more specific to the way in which the user interfaces with the controls of the AGV system. There are many ways to approach the user interface with the AGV system.
Onboard AGV Controls – For many standalone AGV systems, particularly tow type systems, the user directs the AGVs via an onboard keypad terminal. This is a very flexible and low cost method, which allows users to select the AGV destinations and respond to varying conditions.
If your system is relatively simple and/or people load or unload AGVs (by hand, with forklift, etc.), then you may only need to use onboard AGV dispatching to operate your system.
Central Control with Distributed Operator Terminals – Systems that incorporate some level of automation with respect to vehicle dispatching or utilize automatic load transfer often employ a centralized control system. These systems have less need for operators to manage the dispatching of AGVs.
Typically, a central controller (usually a PC) manages the AGV dispatching in response to ‘AGV call and dispatching’ requests that are input via operator terminals distributed throughout the system. For example, an operator may place a load on a stand and call for an AGV to pickup the load via a terminal. They may also input a destination for the load. Operators may also use terminal to request a specific product be delivered to their location by an AGV.
Central Control Schemes with Distributed Electronic Interfaces – This approach is also used for AGV systems that employ a level of automated AGV system management and/or automated load transfer. However, rather than employing remote operator terminals, these systems utilize electronic interfaces to initiate AGV calling and dispatching via the central controller.
The AGV central controller receives inputs from these distributed electronic interfaces, which may include photocells, limit switches, PLCs. other PCs or higher level plant computer systems. For example:
A photocell may be placed at the end of an AGV pickup conveyor to signal when a load is present and ready for transport.
A PLC controlling a stretchwrapper or palletizer would send a message to the central AGV controller when a load is ready for pickup.
Or, a plant production control system may send a message to the AGV system that product should be transported from a given location to another location.
In each case, the AGV central controller takes the inputs and manages the AGV dispatching to accomplish the desired task.
Upgrading Controls Approaches – It is not necessary to start with the most automated level of control for a new AGV project. Typically, you can upgrade the level of control at any time. If you are unsure whether a more sophisticated controls approach is warranted or even desirable, specify a simpler scheme to start out with. Also, it is possible to combine approaches in the same system.
Taking Care Not To Over-Specify – When you over-specify you may unnecessarily limit your choices from different AGV companies. Avoid specifying a brand of device or specific type of display or control, etc. Concentrate on functionality desired as this is often achieved differently by different AGV companies.
Over-specifying will cause some AGV companies to considerable increase their price to meet your requirements. Unfortunately, this biases the comparison between competitive proposals and may not be in your best interest because it may mask other weaknesses in an already compliant company.
Comparisons in Financial Strength and Technology
AGVS comparisons will yield different results depending on the application. Each AGV company possesses a different set of products, skills and general profile. One company may be a better choice for a given application and a poorer choice for a different application.
Financial – An easy place to begin comparing AGV companies is based on financial strength. The importance of this issue directly relates to the size or complexity of the project under consideration.
If the project is simple and generally less than $500,000, nearly every AGV supplier is financially qualified to be considered. However, AGV suppliers have greatly varying financial resources and business size. Consider the size of the project, the length of the project and project risk when weighing the relative financial strength of different AGV suppliers.
Technology – Most larger AGV companies use their own in-house developed technology. Smaller companies tend to buy the technology in a prepackaged form. There is usually no clear-cut, win/loose situation when it comes to comparing technologies. The best approach is to determine if one technology is a better fit with your application needs or if it really does not matter.
Certainly no two AGV projects or environments are the same. No single AGV technology is best for all situations. There are different AGV navigation technologies, different control system capabilities, and different mechanical designs. Comparisons in each area will identify potential strengths and weaknesses for your environment.
A good way to evaluate these factors is to ask each AGV company to answer the same set of questions. An ‘AGVS Vendor Comparison’ document is included to assist you with these issues.
Who Should Go?
Unless you are pretty familiar with AGVS, it should be a high priority to visit some sites. These do not necessarily have to be close replicas of your application, but they should be representative of the general mechanical and controls complexity that is being offered by that AGV supplier for your application.
It is quite important to involve future system users and maintenance people in the evaluation and planning of an AGV system. Therefore, these people need to be part of site visits. They bring a unique perspective on the issues that differentiate potential AGV suppliers. Further, they will be the people that have to live with the system and as such their concerns are best addressed as early as possible.
You can choose to visit a site with or without the AGV vendor to accompany you. There are pros and cons either way. Much of the decision depends on your comfort level with the AGV vendor. He can be valuable in explaining details of the AGV technology and providing direct response to issues that come up during the visit. But, you also don’t want the site’s user feedback to be influenced by the vendor’s presence. If you are comfortable that won’t be the case, then having the vendor present is usually quite helpful.
Questions to Ask and What to Observe
Most important is whom you talk to when you visit a site. Try to speak with the maintenance people and the system operators. A higher level manager does not usually live with the system day in and day out like the floor operators or service people, and can not give you detailed answers.
When you visit one or more sites you should try to ask the same types of questions to be able to compare vendors performance. These questions can focus on several areas of interest.
- How reliable have the vehicles been?
- Do you have a comprehensive service manual with parts list?
- What was your opinion of the maintenance training you received?
- What has been your experience with troubleshooting support from your vendor?
- What has been your experience with parts availability?
- How much maintenance do you have to do to support the AGV system?
- Are the users of the AGV system able to easily interact with it without causing operational problems? If not, what kinds of operator/system problems occur?
- What was the initial user acceptance attitude? How did it progress?
- Is the system being used differently than as originally planned? If so, why?
- What mechanical parts have been replaced and was this due to wear, damage or failure?
- Have you found that the mechanical design of the vehicle is hard or easy to service?
- What is your impression regarding the robustness of the equipment design?
- How often do you have to make mechanical adjustments and to what areas?
- What are the mechanical areas that could be improved and how?
- What electronic parts have been replaced and was this due to wear, damage or failure?
- Have you found that the controls design and layout is hard or easy to service?
- Is there adequate documentation to troubleshoot and service the AGV control systems?
- What is your opinion of the user friendliness of the control system?
- Have you found any serious limitations to the control system’s capabilities (expansion, change, additional functionality)?
- What are your most liked aspects of the controls and what are the most disliked and why?
- Was the documentation provided for the operational level of software adequate?
- How user friendly is the system software? Are there few or many rules for operators to learn?
- Have there been software bugs? If so, what were they and how long did it take to resolve them?
- Is it easy or hard to recover from a system software problem?
- Have you found that your system software is highly custom? If so, has there been any vendor support issues?
What to Observe
- How well does AGV traffic co-exist with other aisle traffic?
- Visit the maintenance shop and note how it is laid out for AGV service.
- Are all areas serviced by the AGV system actually being utilized? If not, why?
- Note quality of installation. Are controls well protected?
- How many hours are on each AGV? (Note AGV hour meter)
- Note movement/flow of AGVs. How efficient is the traffic coordination?
- Have the AGVs been hit? If so, how well did they survive?
- Are there any unnecessarily exposed or vulnerable parts on the AGV?
- Is this a highly standard AGV design used by other customers or is it a fairly unique design?