The top row shows the status quo. Two or three display computers (a.k.a. terminals) are normal. Even four display computers are not unusual. The bottom row shows the cabin of the future with four display computers. The display computers at the left and right beam show the videos from the wing mirror cameras. The display computer in the middle is the instrument cluster. The small display computer on the right-hand side shows additional information. It would be bigger in a harvester.
Let us assume for simplicity that the cabin is equipped with two display computers, which have identical hardware and nearly identical software – except for the main application. When sourced from low-volume, high-customisation electronics manufacturing services (LVHC-EMS), the two display computers, the telematics unit and a switch will cost roughly 2750 Euros for 200 machines.
We will bring down the costs to 2350 Euros, if we replace the two display computers by one computer with two displays. The computer doesn’t have a display and the displays don’t have a computer inside. The telematics unit is folded into an M.2 or PCIe card, which is plugged into a slot of the computer.
We will reduce the costs by another 200 Euros to 2150 Euros, if we source the components from a value-added reseller (VAR) instead of an LVHC-EMS. VARs buy standard components from high-volume low-customisation EMSs (HVLC-EMS). They charge lower prices because of the higher volumes, but they allow only minimal customisations.
For 200 machines, the solution with one computer and two displays saves us 80,000 Euros and 120,000 Euros per year, respectively, when sourced from an LVHC-EMS and from a VAR. We save even more, if we replace more display computers by one computer with displays and if the display computers are more expensive like ISOBUS terminals. We do not only save costs on hardware but also on software, because we consolidate multiple diverse systems into one system.
A clever change of the system architecture leads to sizeable cost savings. It also leads to a system that can easily be extended in many directions – a competitive advantage.
Multiple Display Computers
The diagram above shows a typical harvester or tractor cabin with two display computers. Display Computer 2 is the main display computer, which allows the driver to control and optimise the harvesting process. It is typically attached to the armrest.
Display Computer 1 shows additional information that doesn’t fit on the main display computer and isn’t required all the time. The driver doesn’t interact much with Display Computer 1. He might not interact with it at all. Display Computer 1 may be attached to the left front beam of the cabin and show multiple video streams from cameras at the back or side of the machine. It can also be attached to the right front beam just above the main display computer.
Both display computers are connected to all the CAN buses of the machine. Two to three CAN buses are normal. The display computers show information from the electronic control units (ECUs) like engine speed, engine load, cutting drum speed and dozens more. The driver can change hundreds of ECU parameters like cutting height, unloading conveyor speed and steering mode through the HMI of the display computers.
The Telematics Unit contains a 3G or LTE modem and is the machine’s gateway to the cloud. It also manages the local area network (LAN) consisting of two display computers, two switches and two video cameras – all connected by M12 Ethernet cables. If the telematics unit had a long enough external antenna, it could be placed anywhere in the cabin. This change would eliminate the need for Switch 1.
The telematics unit is also connected to the CAN buses. It forwards the most important machine data from the ECUs to the cloud. Farmers view this data on their phones, tablets or PCs. They see where their machines are, what the yield and diesel consumption are and when the next maintenance is due. Machine manufacturers use the data for remote diagnostics and for insights into the usage of their machines.
Switch 1 is a normal 4-way Ethernet switch, whereas Switch 2 is a 4-way switch with two Power-over-Ethernet (PoE) ports to power the video cameras. All the LAN cables have M12 connectors and coating.
The two display computers are equipped with a quad-core i.MX6 SoC, 1 GB RAM, 4 GB flash, 12-inch display, 1280×800 resolution and capacitive touch. They use Deutsch connectors for power, CAN, USB and RS232 and an M12 connector for LAN. Display computers with such a specification were ubiquitous at Agritechnica 2017 as I established in my trip report Agritechnica 2017: What’s New for Terminals.
One Computer with Multiple Displays
The idea is to have one central computer, which is typically contained in a Silver Box, with two displays – instead of two display computers. The Silver Box does not have a display. The two touch displays – Display 1 and Display 2 – don’t have a computer inside. They are connected with the Silver Box over display port (DP), LVDS, HDMI or even VGA.
The Silver Box contains the computer part and Display 1 and Display 2 the display parts of the orgininal display computers. Otherwise, the hardware specification is the same as outlined in the section Multiple Display Computers.
Switch 1 is not needed any more, because the two LAN cables to the display computers have been replaced by display cables. The Telematics Unit becomes a part of the Silver Box as well. The LTE modem and the GPS receiver can be integrated into the main board of the Silver Box or put on a PCIe or M.2 card, which is plugged into the slots of the main board. This eliminates the need for a separate telematics unit and for the LAN cable between Switch 1 and the Telematics Unit.
The idea to have one computer with multiple displays is not new. It is fairly common in premium cars and SUVs. A powerful computer in a Silver Box drives up to four displays: the instrument cluster, the infotainment system, the rear-seat entertainment system and the passenger entertainment system. To my knowledge, manufacturers of harvesting machines have not picked up this idea yet.
Pros and Cons
Let us look at the pros and cons of the approach with one computer and multiple displays.
Pro: Hardware Costs
The following table compares the hardware costs for three variants.
- The second column shows the costs for two display computers sourced from a low-volume high-customisation electronics manufacturing service (LVHC-EMS, for short).
- The third column shows the costs for one silver box with two displays sourced form an LVHC-EMS.
- The fourth column shows the costs for one silver box with two displays sourced from a value-added reseller (VAR).
I explain the terms LVHC-EMS and VAR in the appendix. The prices are in Euro for 200 units per year. They are estimates based on what I learnt from several projects.
When sourced from an LVHC-EMS, the solution with one computer and two displays is 400 Euros or 15% cheaper than the solution with two display computers. When sourced from a VAR, it is even 600 Euros or 22% cheaper. The cost savings for 200 units are 80,000 and 120,000 Euros, when sourced from an LVHC-EMS and from a VAR, respectively!
The VAR solution becomes even cheaper, if we are willing to compromise on the robustness of the connectors (e.g., RJ45 instead of M12; M12, USB or D-SUB instead of Deutsch) and the protection level. The LVHC-EMS solutions become more expensive than the VAR solution, if we want a customised solution that differs considerably from the “standard” model manufactured by the LVHC-EMS. Requiring a Renesas R-Car M2 SoC instead of an NXP i.MX6 SoC, four instead of two CAN connectors or onboard Wifi, Bluetooth, LTE and GPS would be such considerable differences.
The cost savings of the silver-box solution are even more striking, if we replace three or more display computers or more expensive display computers by displays. An example for more expensive display computers are ISOBUS terminals. They easily cost 700 Euros more than their non-ISOBUS cousins.
Of course, we could be pleased with the cost savings and use the silver-box solution as described. Or, we could be smart and invest some of the cost savings in a more powerful microprocessor. For example, we could replace the i.MX6 by the more powerful i.MX8 by spending roughly 80 Euros of our savings.
Pro: Consolidating Multiple Systems into a Single System
We have assumed so far that the two displays computers have the same hardware and the same operating system. They only differ in the single application running on them. This is the rare exception. In reality, we must deal with a zoo of heterogenous display computers, as we can see in Figure 1.
- The display computers in a cabin run on different hardware. Some run on Intel microprocessors, some on ARM microprocessors. Some still run on microcontrollers.
- Some display computers have ISOBUS connectors, some not.
- The display computers run on different operating systems like Linux, Windows and several RTOSs. Even if display computers run on the “same” operating system, they run on different versions and configurations.
- The display computers use different window managers. Even on Linux, they may use different window managers: Qt compositor based on Wayland, X11, single window (eglfs) or directfb.
- The applications, which run on the display computers, use different GUI and application libraries.
- Some display computers update their software over the air, others over a CAN or LAN connection with a laptop. The update cycles differ wildly.
- And the list goes on …
This variety adds considerable complexity and costs to our development projects. We need internal or external developers with very different skill sets. Developers, who have all the above skills, are rare and expensive. So, we must hire more developers and train them in all these skills. Developers will end up doing things multiple times.
The situation will improve dramatically, if we use one computer in a silver box with multiple displays. The silver-box computer is powered by one microprocessor (typically: an ARM-based SoC like NXP i.MX6), has one type of connectors for machine communication (typically: ISOBUS) and runs on one operating system (typically: Linux custom-built with Yocto). All applications use the same GUI and application framework (typically: Qt) and are displayed by the same window manager (typically: a Wayland-based Qt compositor). The Linux system and the applications are updated through one update mechanism (typically: Linux package manager).
Using a single computer lets us get away with a smaller team of developers, whose skill set is more focused and more widely available. It implies significantly lower software development costs or faster development and update cycles.
Con: Hardware Redundancy
We assume for this section that the original display computers have identical hardware and nearly identical software. If the main display computer, Display Computer 2, fails during the harvest, drivers can remove the main display computer and install the other display computer, Display Computer 1, in its place. They must unfasten and fasten a couple of screws, pull out and plug in some cables, and reboot the display computer to the correct main application. After less than ten minutes, they are good to continue with the harvest.
Traditionally, the replacement of the broken display computer would take at least a day. The harvester would stand still, until the replacement arrives from the harvester manufacturer. Such a delay is the worst and most expensive thing during a harvest. Drivers often work in shifts to keep the harvesters running 24/7 during the short harvest season. Hence, having a “redundant” display computer at hand is a big boon to the customers.
The silver-box solution benefits from hardware redundancy as well. If the main display, Display 2, breaks, drivers can replace it with the other display, Display 1, and continue with the harvest after some minutes. If display computers break, displays are the culprit in more than 80% of the cases. Manufacturers could cover the remaining 20% by providing a spare silver box for every fifth machine. All in all, hardware redundancy isn’t much of a con for the silver-box solution.
My thanks go to Romeo Stübling from SPHINX Computer. He made me aware of the sizeable cost savings, when components for the silver-box solution are sourced from a value-added reseller (VAR). His employer, SPHINX Computer, is a VAR that offers great additional value.
Appendix: What are HVLC-EMSs, LVHC-EMSs and VARs?
Component prices depend heavily on the source, from which you buy them. Three sources are relevant for this post: high-volume low-customisation electronics manufacturing services (HVLC-EMS) , low-volume high-customisation electronics manufacturing services (LVHC-EMS) and value-added resellers (VAR).
EMS assemble the printed circuit board (PCB) according to your requirements. They add parts like CAN connectors, LTE modems, GPS receivers or Wifi/BT to the board. They decide on the touch and display type. They build the enclosures and make them conform with different protection levels like IP65 and IP66. In short, an EMS provider customises the display computer, the silver box and the displays to your needs.
EMS companies come in two varieties: high-volume low-customisation EMS (HVLC-EMS) and low-volume high-customisation EMS (LVHC-EMS). Components sourced from a HVLC-EMS cost less than components sourced from a LVHC-EMS, because HVLC-EMS produce components in much higher volumes than LVHC-EMS. However, you must take whatever the HVLC-EMS offer. Most likely, you will get neither Deutsch connectors for CAN, USB, RS232 and power nor M12 connectors for Ethernet. You won’t get a customised enclosure either.
VARs buy the enclosures, the fully assembled boards and the touch monitor from a HVLC-EMS. VARs assemble these parts, wire up the board with the connectors of the enclosure, and plug a PCIe or M.2 card into the slots on the board. They may even add some connectors to the enclosure and ensure the required protection level. They don’t modify the board or replace the display in the monitor.