preloader

Search Here

INDUSTRIAL AG VEHICLE CONTROL

Project Scope

Two successful application equipment manufacturers formed a joint venture and contracted AVISTA to help build and maintain software for a touch screen in-cab display, a side-fill station micro-display, and an Electronic Control Unit to control the solution system for a new liquid system option for their nutrient applicator floater chassis. AVISTA engineers developed and tested software that is capable of the following:

  • Automatically filling a 2000 gallon solution tank and a 200 gallon rinse tank at up to 300 gallons/minute.
  • Utilizing up to two solution pumps to automatically adjust output for different filling, mixing, and rinsing functions or spraying out a 90-foot boom at different rates based on desired application rate and changing vehicle speed. Spraying rates may be as low as 5 gallons/minute for pre-emergent herbicide and as high as 400 gallons/minute for high-rate, pre-plant liquid fertilizer.
  • Actuating two-way and three-way valves to manage flow between circuits for different functions, which included a configurable automated rinse procedure to flush plumbing components with clean water to prevent corrosion and contamination, mixing and agitating chemicals, pressure washing, and providing water for jug/bowl rinsing at the filling station.
Solution

AVISTA’s engineers ramped up their system knowledge and quickly recognized the opportunity for software re-use by analyzing two existing solution system controllers. A software baseline was created and modifications were made to the models and hand code, a plant model was integrated to begin testing in a simulated test environment, and tests were written for basic controller functionality. Eventually, software was loaded on the test vehicle and basic functionality was verified.

Working closely with system engineers and utilizing Scrum, AVISTA’s team worked to deliver the highest value features first. During two-week sprints, these features were developed, simulated, and tested on actual vehicles. As functionality grew more complex, the plant model was expanded to include tanks, pumps, valves, fluid flows, and nozzles representing the sprayer. Using this closed-loop testing, simulated test results could be corroborated against field test results. Because the early “feasibility-build” vehicles had electronic harnesses assembled by hand, AVISTA had to first troubleshoot and help diagnose electrical, assembly, and hydraulic issues on-site before the software could be verified. Customer feedback praised the AVISTA team for their breadth and depth of understanding and ability to build and test their software solutions.

Strong system experience coupled with analysis of existing software allowed AVISTA engineers to inform System Engineering when discrepancies were found between assumed functionality in the requirements and actual software implementation. AVISTA helped to facilitate resolution and keep the project moving smoothly.
Using the Agile framework, AVISTA’s team was fully integrated into the customer’s software planning and delivery process. As AVISTA’s key stakeholders, Systems Engineers attended sprint reviews where simulated and field tested results of new and existing features were discussed and analyzed. The customer feedback and demonstration of feature functionality enabled rapid software development.
AVISTA’s team members traveled regularly to the customer sites to coordinate and perform early software integration testing on a prototype vehicle before the customer’s official integration test team took over. AVISTA’s team members were able to identify and help resolve build problems with electrical, mechanical, and hydraulic systems so that software integration testing could continue. After the customer began their own field testing, AVISTA analyzed the data from these test runs and drove improvements to the software.
Development of effective system models allowed model-in-the-loop and software-in-the-loop testing of all-new control features that enabled correlation with vehicle testing results, aided in catching errors in the software before they became defects, and drove improvements to the software.
Using an automation server, AVISTA’s software team was able to automate execution of its regression tests, code generation and compilation, and software releases.
Due to AVISTA’s team experience with guidelines for safety-critical software and CMMI best practices, AVISTA was able to deliver value early and with minimal defects. The customer recognized that AVISTA was able to teach them how to perform better software testing and considered AVISTA’s team execution from project start to finish as an example for others to follow.