How can a non-autonomous compaction roller be converted to become temporarily autonomous while it performs a 500hours endurance test? Particularlysince the compaction rollers in question is not built to be autonomous and shall not be autonomous after the endurance test is completed. The autonomous system shall also be adaptable to all compaction rollers which Dynapac is developing and shall be moved to another machine when the endurance test is completed. In this thesis a concept is engineered of how the whole autonomous system will work and a prototype is fabricated of how to convert the current manual mechanical steering to be performed by a computer. The steering prototype has been tested on a Dynapac CC4200 double drumasphalt compaction roller and worked as intended. To develop this, anextensive risk analysis is also established andwith it a requirements list of what's needed to be fulfilled when performing autonomous testing of a compaction roller. The work has been done using the method “design thinking” which is a collection of multiple methods to create new concepts and ideas. The final concept resulted in a navigation system which uses GNSS for path planning and limitation of the operation area. It also uses radar to detect foreign objects in its path to prevent a collision. Multiple systems arealso proposed to be used for malfunction detection of the roller, which is a major part of a human operator’sjob when testing out new machines. The test track for the machine was undefined and also hadto be engineeredas part of the concept. It resultedin closing the area of operation with a mesh fence to prevent access to the area from unauthorised personnel and geo-fence to prevent the machine from escaping. Access to the area is only granted to authorized personnel and only when the autonomous rolleris shut off. Due to the machines in question isn’t fully developed, theycan’t be trusted enough to have people inside the area of operation asthe autonomous machineis operating. / Hur kanen icke-autonom vägvältomvandlas tillatt bli tillfälligt autonom medan den utför ett 500timmar långttidsprov?Särskilt sedanvägvältenifrågainte ärbyggd för att vara autonom och ska intevara autonom efter attlångtidsprovetär slutfört. Det autonoma systemet skaävenkunna anpassas tillalla vältar som Dynapac utvecklar och ska flyttas till en annan maskin närlångtidsprovetär klart. Idenna avhandling konstrueras ett koncept för hur hela det autonoma systemet kommer att fungera ochenprototyp tillverkaspå hur man konverterar den nuvarande manuella mekaniska styrningen till attstyras av en dator. Styrprototypen testades på en Dynapac CC4200 asfaltsvält med dubbla valsar ochfungerade bra. En omfattande riskanalys utvecklades ochlika såen kravlista över vad som behöveruppnås vid autonom testning av en vägvält. Arbetet har gjorts med hjälp av metoden “designthinking”, vilket är en samling av flera metoder för att skapa nya koncept och idéer. Det slutgiltigakonceptet resulterade i ett navigationssystem som använder GNSS för navigering och begränsning avkörområdet. Den använder också radar för att upptäcka främmande föremål i sin vägvilketförhindrarkollision. Flera system föreslås användasförfunktionsfelsdetektering på välten, vilket är en viktig delav en mänskligoperatörs arbetevid provning av nya maskiner. Maskinen kommer att vara i ett slutetområde som är avskilt med ett nätstängsel.Tillträde till området ges endast till behörig personal ochendast när den autonoma välten är avstängd. På grund avmaskinerna ifråga inte är fullt utvecklade,kan de inte litas på tillräckligt för att ha personer inom körområdet medan det autonoma systemet är idrift.

Symptoms of harsh environments create pressure on autonomous machines to work wellwith the use of measurement devices. Sensors used within construction equipment vehiclessuffer from being covered with fogs of dirt, oil, water, and dust. Autonomous vehicle’sdependency on fully functional sensors paves roads for research to be made on ways to keepsensor data from being compromised and automation to be improved.In cooperation with Dynapac Compaction Equipment AB, this master’s thesis problemopted for innovative and sustainable product development to tackle the problem. The thesiscomprises of 30 credits. In this thesis, work will be done emphasizing on ways to keepautonomous performance at peak for construction equipment vehicles with low to no capacityfor sophisticated equipment and supply of power and water. For reference, the project targetsasphalt rollers due to their relevance at Dynapac.With the use of innovative product development tools, the problem is to be tackled with anengineering approach revolving around the design thinking process. All progress will beconstituted at Dynapac’s facilities in Karlskrona. Interviews and dialogue with relevantindividuals are to be held to create value based on customer needs which has been of highpriority and is to be iteratively monitored throughout the project. Prototyping will play a partwith large amount of testing to prove concept, including processes of 3dmodeling andmanufacturing of components.With measurement incorrections reaching up to 80 degrees out of a measurement spanwithin the range of 0 to 150 degrees, three concepts have been established, each of which areequipment neutral to benefit construction vehicles which has shown to be valuable within theindustry.A pipe concept which’s purpose is to hide the sensor further and create distance betweensensor lens and measurement surface has been established. The concept is feasible for theconstruction industry, protects the sensor from debris and supports autonomous systems.