Ant colony algorithm (ACA) is employed and improved in routing protocol of information acquisition of manufacturing process based on wireless networks. Weighting factor for paths selection is defined as an exponential variable in adaptive ACA. In this way, it prevents the weighting factor from excess increasing or rapid reducing to 0 which results in local optimum. This approach can dynamically adjust paths selection and improve global search ability by optimizing global policy. Adaptive ACA consumes the least time in the process of searching the most optimized paths and searches the shortest paths under the same of iterative loops. Under the same condition of the information heuristic factor and the expected heuristic factor, the algorithm shows good adaptation, realizes the load balancing between paths and resolves the dynamic adjustment problem.